{
  "id": 461058,
  "title": "207th Place Solution for the Open Problems – Single-Cell Perturbations.  Late submission 0.576 private LB.",
  "url": "/competitions/open-problems-single-cell-perturbations/writeups/liuda-cheldieva-207th-place-solution-for-the-open-",
  "author_name": "",
  "post_date": "2023-12-19T00:00:53.907Z",
  "votes": 6,
  "comment_count": 1,
  "views": 0,
  "content": "<h2>1. Introduction</h2>\n<p>I want to thank Kaggle and everyone involved for hosting such an intriguing competition.</p>\n<p>Two topics are what I describe.<br>\n1.1.   late submission<br>\n1.2.   solution </p>\n<p>It seems that one of the models is better for one compound but weak for other compound.</p>\n<p>The question remains: how to choose the right drug model for compound?</p>\n<p>Considering that cells of various types have different models.</p>\n<p>Could you please clarify why a model is good for a compound?<br>\nWhy is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?</p>\n<p>Initially, it is necessary to cover a larger portion of the error with models. <br>\nWith an error of 0.661 and without knowing which compounds have large errors, it is difficult to say whether the model is suitable for the compound.</p>\n<h2>2. Context</h2>\n<p>Business context: <a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/overview\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/overview</a><br>\nData context: <a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/data\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/data</a></p>\n<h2>3. Overview of the approach (  late submission)</h2>\n<p>I like to discover which model is better for compounds.<br>\nThe result is better if you select one model for each drug and cell type.<br>\nIt's possible that there's a chemical explanation for model selection.</p>\n<p>We lack the skill to select from all the models available.<br>\nAll models have been added.</p>\n<table>\n<thead>\n<tr>\n<th>bland</th>\n<th>Private score</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Weight sum  of 10 models</td>\n<td>0.718</td>\n</tr>\n<tr>\n<td>Select one model for each compound</td>\n<td>0.657</td>\n</tr>\n<tr>\n<td>Multiply the compound by a constant (between 1.4 and 3.09)</td>\n<td>0.576</td>\n</tr>\n</tbody>\n</table>\n<h3>Late submission 0.577:</h3>\n<table>\n<thead>\n<tr>\n<th>Submition id</th>\n<th>Model</th>\n<th>Multiplecator is Post Processing</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>0</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>1</td>\n<td>2nd place result_11  776 575 *50% + + 50% train  <strong>B cells, Alvocidib</strong></td>\n<td></td>\n</tr>\n<tr>\n<td>2</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>3</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>4</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>5</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>6</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>7</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>8</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>9</td>\n<td>24plase LS NK  784</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>10</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>11</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td></td>\n</tr>\n<tr>\n<td>12</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td></td>\n</tr>\n<tr>\n<td>13</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>14</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>15</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>16</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>17</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>18</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>19</td>\n<td>1plase 729 569</td>\n<td>3.09</td>\n</tr>\n<tr>\n<td>20</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>21</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>22</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>23</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>24</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>25</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>26</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>27</td>\n<td>2nd place result_11  776 575</td>\n<td></td>\n</tr>\n<tr>\n<td>28</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td></td>\n</tr>\n<tr>\n<td>29</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>30</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>31</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>32</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>33</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>34</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>35</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>36</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>37</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>38</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>39</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>40</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>41</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>42</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>43</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>44</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>2</td>\n</tr>\n<tr>\n<td>45</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>46</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>47</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>48</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>49</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>50</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>51</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>52</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>53</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>54</td>\n<td>24plase LS NK  784</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>55</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>56</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>57</td>\n<td>24plase LS RG  834</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>58</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>59</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>60</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>61</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>62</td>\n<td>1plase 729 569</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>63</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>64</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>65</td>\n<td>24plase LS NK  784</td>\n<td>2.9</td>\n</tr>\n<tr>\n<td>66</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>67</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>68</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>69</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>70</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>71</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>72</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>73</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>74</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>75</td>\n<td>24plase LS RG  834</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>76</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>77</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>78</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>79</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>80</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>81</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>82</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>83</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>84</td>\n<td>24plase LS NK  784</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>85</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>86</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>87</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>88</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>89</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>90</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>91</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>92</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>93</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>94</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>95</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>96</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>97</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>98</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>99</td>\n<td>2nd place result_15  758 551</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>100</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>101</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>102</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>103</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>104</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>105</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>106</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>107</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>108</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>109</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>110</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>111</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>112</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>113</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>114</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>115</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>116</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>117</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>118</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>119</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>120</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>121</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>122</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>123</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>124</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>125</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>126</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>127</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>128</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>129</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>130</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>131</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>132</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>133</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.675</td>\n</tr>\n<tr>\n<td>134</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>135</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>136</td>\n<td>24plase LS NK  784</td>\n<td>2.031</td>\n</tr>\n<tr>\n<td>137</td>\n<td>2nd place result_15  758 551</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>138</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>139</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>140</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>141</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>142</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>143</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>144</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>145</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>146</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>147</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.21</td>\n</tr>\n<tr>\n<td>148</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>149</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>150</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td></td>\n</tr>\n<tr>\n<td>151</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>152</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>153</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>154</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>155</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>156</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>157</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td></td>\n</tr>\n<tr>\n<td>158</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td></td>\n</tr>\n<tr>\n<td>159</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>160</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>161</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>162</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>163</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>164</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>165</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.65</td>\n</tr>\n<tr>\n<td>166</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>167</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>168</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>169</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>170</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>171</td>\n<td>pyBoost 760. 586</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>172</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>173</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>174</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>175</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>176</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>177</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>178</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>179</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>180</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>181</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>182</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>183</td>\n<td>24plase LS RG  834</td>\n<td>1.5</td>\n</tr>\n<tr>\n<td>184</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>185</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>186</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>187</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>188</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>189</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>190</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td></td>\n</tr>\n<tr>\n<td>191</td>\n<td>24plase LS NK  784</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>192</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>193</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>194</td>\n<td>2nd place result_15  758 551</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>195</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>196</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>197</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>198</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>199</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>200</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>201</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>202</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>203</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td></td>\n</tr>\n<tr>\n<td>204</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>205</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>206</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>207</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>208</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>209</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>210</td>\n<td>24plase LS RG  834</td>\n<td>1.870</td>\n</tr>\n<tr>\n<td>211</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>212</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>213</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>214</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>215</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>216</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>217</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>218</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>219</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>220</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>221</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>222</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>223</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>224</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>225</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>226</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>227</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>228</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>229</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>230</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>231</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>232</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>233</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>234</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>235</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>236</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>237</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>238</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>239</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.351</td>\n</tr>\n<tr>\n<td>240</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>241</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>242</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>243</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>244</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>245</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>246</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>247</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>248</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td></td>\n</tr>\n<tr>\n<td>249</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>250</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>251</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>252</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>253</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>254</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>255</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n</tbody>\n</table>\n<h2>4. Data preprocessing (  solution)</h2>\n<p>The models employed are NN and pyboost.</p>\n<p>Cross validation is not used</p>\n<h2>5. Method (  late submission)</h2>\n<h2>Ensembling <strong>late submissions</strong></h2>\n<p>I evaluate for various compounds the models:</p>\n<table>\n<thead>\n<tr>\n<th>model</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>1th place</td>\n</tr>\n<tr>\n<td>2nd place</td>\n</tr>\n<tr>\n<td>3rd place</td>\n</tr>\n<tr>\n<td>4th place</td>\n</tr>\n<tr>\n<td>24th place NK</td>\n</tr>\n<tr>\n<td>24th place RG</td>\n</tr>\n<tr>\n<td>pyBoost</td>\n</tr>\n</tbody>\n</table>\n<p>Green select indicates that the model has the lowest error among all.</p>\n<table>\n<thead>\n<tr>\n<th>late submissions</th>\n<th>private LB</th>\n<th>public LB</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>1th place+2nd place+3rd place+4th place+24th place NK+24th place RG+pyBoost+50%+mult</td>\n<td><strong>0,597</strong></td>\n<td>0,548</td>\n</tr>\n</tbody>\n</table>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F1dadc65287d17a1469ebbd71e0e95364%2F2023-12-13%20%2023.15.41.png?generation=1702499882391676&amp;alt=media\" alt=\"\"></p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F887cf9a7156a8a35b83850cdf7168c8f%2F2023-12-13%20%2023.16.16.png?generation=1702499901312868&amp;alt=media\" alt=\"\"></p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F8600e331c1e6304df98f7bcaab85a41e%2F2023-12-13%20%2023.16.44.png?generation=1702499920405715&amp;alt=media\" alt=\"\"></p>\n<p>Why is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?</p>\n<h2>6. Final result (  solution,   late submission)</h2>\n<table>\n<thead>\n<tr>\n<th>solution</th>\n<th>private LB</th>\n<th>public LB</th>\n<th>solution or late submission</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>bland</td>\n<td><strong>0,577</strong></td>\n<td>0,53</td>\n<td>late submission</td>\n</tr>\n<tr>\n<td>Pyboost max_depth 10, ntrees 5000, lr 001, colsample 035</td>\n<td>0,763</td>\n<td>0,586</td>\n<td>solution</td>\n</tr>\n<tr>\n<td>Pyboost max_depth 12, ntrees 5000, lr 001, colsample 075,  n_components 49,  T8 b7t17</td>\n<td><strong>0,760</strong></td>\n<td>0,586</td>\n<td>solution</td>\n</tr>\n<tr>\n<td>Pyboost max_depth 12, ntrees 1000, lr 001, colsample 035, n_components 50,   T8 b7t1 bcell7 mie6</td>\n<td>0,761</td>\n<td>0,578</td>\n<td>solution</td>\n</tr>\n<tr>\n<td>quantile(0.7)</td>\n<td>1,003</td>\n<td>0,797</td>\n<td>solution</td>\n</tr>\n</tbody>\n</table>\n<h2>What help (  solution)</h2>\n<p><strong>Drop T8 cell</strong><br>\n<strong>Repeat 17 drug   - 7 time</strong>    +0,001 to Pyboost</p>\n<h2>What help (  late submission)</h2>\n<p>The main score improvement is in <strong>4th Place Magic Postprocessing</strong>  - <br>\n<strong>multiply the compound by a constant</strong> (between 1.4 and 3.09).</p>\n<p>Base score 3rd place submission Private LB 732 , public LB 548</p>\n<table>\n<thead>\n<tr>\n<th>Algorithm</th>\n<th>Private LB  improve</th>\n<th>Private LB</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Base   3rd place submission</td>\n<td></td>\n<td>0.732</td>\n</tr>\n<tr>\n<td>select model from 10 for compound</td>\n<td>+0.075</td>\n<td>0.657</td>\n</tr>\n<tr>\n<td><strong>B cells, ABT-199</strong> =    50% model  <strong>B cells, ABT-199</strong>  + 50% train  <strong>B cells, Alvocidib</strong></td>\n<td>+0.015</td>\n<td></td>\n</tr>\n<tr>\n<td>Add constant (+0.1, -0.1, +3 B cells  BMS-387032)</td>\n<td>+0.008</td>\n<td></td>\n</tr>\n<tr>\n<td>50% model +  50% compound</td>\n<td>+0.008</td>\n<td>0.626</td>\n</tr>\n<tr>\n<td>4th Place Magic Postprocessing of 1 compound: <strong>B cells, BMS-387032</strong>  MULT 3.09</td>\n<td>+0.027</td>\n<td>0.609</td>\n</tr>\n<tr>\n<td>4th Place Magic Postprocessing of 55 compounds MULT 1.2, 1.3, 2</td>\n<td>+0.009</td>\n<td><strong>0.600</strong></td>\n</tr>\n</tbody>\n</table>\n<h2>+0.008 Private LB detail:</h2>\n<p><strong>Compound A</strong> = <strong>Compound A</strong> model * 50% + <strong>Compound B</strong> train * 50%</p>\n<table>\n<thead>\n<tr>\n<th>50% model  submission Cell, Compound</th>\n<th>50% train Cell, Compound</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>27 B cells, CGM-097</td>\n<td>train row 3. T regulatory cells,    Clotrimazole</td>\n</tr>\n<tr>\n<td>65 B cells, Lamivudine</td>\n<td>train row 7. T regulatory cells,    Mometasone Furoate</td>\n</tr>\n<tr>\n<td>147 Myeloid cells, BMS-536924</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>152 Myeloid cells,CC-401</td>\n<td>train row 1. T cells CD4+,  Clotrimazole</td>\n</tr>\n<tr>\n<td>155 Myeloid cells,CGM-097</td>\n<td>train row 0. NK cells,  Clotrimazole</td>\n</tr>\n<tr>\n<td>157 Myeloid cells,Canertinib</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>207    Myeloid cells,PF-03814735</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>214    Myeloid cells,Prednisolone</td>\n<td>train row 6. T cells CD8+,  Mometasone Furoate</td>\n</tr>\n<tr>\n<td>229    Myeloid cells,SLx-2119</td>\n<td>train row 2. T cells CD8+,  Clotrimazole</td>\n</tr>\n<tr>\n<td>241    Myeloid cells,Tacalcitol</td>\n<td>train row 234. Myeloid cells,   Porcn Inhibitor III</td>\n</tr>\n<tr>\n<td>247    Myeloid cells,Tosedostat</td>\n<td>train row 1. T cells CD4+,  Clotrimazole</td>\n</tr>\n<tr>\n<td>248    Myeloid cells,Trametinib</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>251    Myeloid cells,Vanoxerine</td>\n<td>train row 1. T cells CD4+,  Clotrimazole</td>\n</tr>\n</tbody>\n</table>\n<h2>7. Sources (  solution)</h2>\n<p><a href=\"https://www.kaggle.com/code/liudacheldieva/blend-for-single-cell-perturbations-dd4fc8\" target=\"_blank\">https://www.kaggle.com/code/liudacheldieva/blend-for-single-cell-perturbations-dd4fc8</a></p>\n<p><strong>0.760</strong> <a href=\"https://www.kaggle.com/code/jeffreylihkust/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/notebook?scriptVersionId=150765833\" target=\"_blank\">https://www.kaggle.com/code/jeffreylihkust/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/notebook?scriptVersionId=150765833</a><br>\n0.761 <a href=\"https://www.kaggle.com/code/emmawilsonev/copy-of-pyboost-secret-grandmaster-s-to-1d68b4?scriptVersionId=150583811\" target=\"_blank\">https://www.kaggle.com/code/emmawilsonev/copy-of-pyboost-secret-grandmaster-s-to-1d68b4?scriptVersionId=150583811</a><br>\n<a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/458750\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/458750</a><br>\n<a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/460191\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/460191</a></p>\n<h2>8. Sources (  late submission)</h2>\n<p>Private score <strong>0.577</strong>  <a href=\"https://www.kaggle.com/code/liudacheldieva/fork-of-ensembling\" target=\"_blank\">https://www.kaggle.com/code/liudacheldieva/fork-of-ensembling</a> <br>\nsummation.csv 0.577: <a href=\"https://www.kaggle.com/datasets/liudacheldieva/op-0-670\" target=\"_blank\">https://www.kaggle.com/datasets/liudacheldieva/op-0-670</a> </p>\n<h2>9. Used submission.csv (  late submission)</h2>\n<p>import_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')<br>\nimport_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')<br>\nimport_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')<br>\nimport_760pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')<br>\nn1_776_result_11  = pd.read_csv('/kaggle/input/2-pla-submit/result (11).csv', index_col='id')<br>\nn5_760_submission_pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')<br>\nn6_776_submission_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')<br>\nn7_792_submission_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')<br>\nn8_813_submission_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')<br>\nn3_750_result_9 = pd.read_csv('/kaggle/input/2-pla-submit/result (9).csv', index_col='id')<br>\nn4_767_result_8 = pd.read_csv('/kaggle/input/2-pla-submit/result (8).csv', index_col='id')<br>\nn2_758_result_15 = pd.read_csv('/kaggle/input/2-pla-submit/result (15).csv', index_col='id')<br>\nn9_736_3rd_place_psevdo_submission = pd.read_csv('/kaggle/input/3rd-place/3rd place psevdo submission(14).csv', index_col='id')<br>\nn12_746_OP_pub530_myModels_exp1 = pd.read_csv('/kaggle/input/4-pl-sub-late/OP_pub530_myModels_exp1.csv', index_col='id')<br>\nimport_732 = pd.read_csv('/kaggle/input/3rd-place/3rd place submission.csv', index_col='id')<br>\nimport_746 = pd.read_csv('/kaggle/input/4-pl-sub-late/4th_Place_Ensembling_submission.csv', index_col='id')<br>\nn13_729_1_pl = pd.read_csv('/kaggle/input/submit-best/submission.csv', index_col='id')<br>\nn14_719_sub_b = pd.read_csv('/kaggle/input/1st-place-solution/submission.csv', index_col='id')<br>\nimport_670 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729.csv', index_col='id')<br>\nimport_657 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785.csv', index_col='id')<br>\nimport_626 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add.csv', index_col='id')<br>\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')<br>\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')<br>\nm1_792_4th_submission = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')<br>\nm2_784_LS_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_NK.csv', index_col='id')<br>\nm3_834_LS_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_RG.csv', index_col='id')<br>\nm4_775_LS_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T4.csv', index_col='id')<br>\nm5_959_LS_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T8.csv', index_col='id')<br>\nm6_784_SVD_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_NK.csv', index_col='id')<br>\nm7_834_SVD_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_RG.csv', index_col='id')<br>\nm8_775_SVD_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T4.csv', index_col='id')<br>\nm9_959_SVD_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T8.csv', index_col='id')<br>\nm10_773_SVD_b_nk_m_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_nk_m_t4.csv', index_col='id')<br>\nm11_786_SVD_b_t4_m_nk = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_t4_m_nk.csv', index_col='id')<br>\nm12_768_SVD_nk_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4.csv', index_col='id')<br>\nm13_771_SVD_nk_t4_t8_rg = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4_t8_rg.csv', index_col='id')<br>\nm14_746_op_late_submission = pd.read_csv('/kaggle/input/op-late-sub-pub530-mymodels-exp1/submission.csv', index_col='id')<br>\nm15_863_plase30_submission = pd.read_csv('/kaggle/input/30-op2scp-transformer-deberta-v3-small-demo/submission.csv', index_col='id')<br>\nm17_769_submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50 = pd.read_csv('/kaggle/input/pyboost-u900/submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')<br>\nm18_790_submission_calibrated_max_d = pd.read_csv('/kaggle/input/pyboost-u900/submission_calibrated_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')<br>\nM19_785_13th_place_mlp = pd.read_csv('/kaggle/input/13th-place-solution-mlp/submission_example_MLP.csv', index_col='id')<br>\nimport_755_10pl = pd.read_csv('/kaggle/input/10th-place/sub_submission.csv', index_col='id')<br>\nimport_748_26pl = pd.read_csv('/kaggle/input/op-26-place/submission.csv', index_col='id')<br>\nimport_577 = pd.read_csv('/kaggle/input/op-0-670/submi_mu_0.577.csv', index_col='id')</p>\n<p>Thank you for taking the time to read. Feel free to ask any questions you may have.</p>",
  "messages": [
    {
      "id": "2558901",
      "postDate": "12/12/2023 12:29:18",
      "content": "<h2>1. Introduction</h2>\n<p>I want to thank Kaggle and everyone involved for hosting such an intriguing competition.</p>\n<p>Two topics are what I describe.<br>\n1.1.   late submission<br>\n1.2.   solution </p>\n<p>It seems that one of the models is better for one compound but weak for other compound.</p>\n<p>The question remains: how to choose the right drug model for compound?</p>\n<p>Considering that cells of various types have different models.</p>\n<p>Could you please clarify why a model is good for a compound?<br>\nWhy is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?</p>\n<p>Initially, it is necessary to cover a larger portion of the error with models. <br>\nWith an error of 0.661 and without knowing which compounds have large errors, it is difficult to say whether the model is suitable for the compound.</p>\n<h2>2. Context</h2>\n<p>Business context: <a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/overview\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/overview</a><br>\nData context: <a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/data\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/data</a></p>\n<h2>3. Overview of the approach (  late submission)</h2>\n<p>I like to discover which model is better for compounds.<br>\nThe result is better if you select one model for each drug and cell type.<br>\nIt's possible that there's a chemical explanation for model selection.</p>\n<p>We lack the skill to select from all the models available.<br>\nAll models have been added.</p>\n<table>\n<thead>\n<tr>\n<th>bland</th>\n<th>Private score</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Weight sum  of 10 models</td>\n<td>0.718</td>\n</tr>\n<tr>\n<td>Select one model for each compound</td>\n<td>0.657</td>\n</tr>\n<tr>\n<td>Multiply the compound by a constant (between 1.4 and 3.09)</td>\n<td>0.576</td>\n</tr>\n</tbody>\n</table>\n<h3>Late submission 0.577:</h3>\n<table>\n<thead>\n<tr>\n<th>Submition id</th>\n<th>Model</th>\n<th>Multiplecator is Post Processing</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>0</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>1</td>\n<td>2nd place result_11  776 575 *50% + + 50% train  <strong>B cells, Alvocidib</strong></td>\n<td></td>\n</tr>\n<tr>\n<td>2</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>3</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>4</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>5</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>6</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>7</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>8</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>9</td>\n<td>24plase LS NK  784</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>10</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>11</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td></td>\n</tr>\n<tr>\n<td>12</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td></td>\n</tr>\n<tr>\n<td>13</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>14</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>15</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>16</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>17</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>18</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>19</td>\n<td>1plase 729 569</td>\n<td>3.09</td>\n</tr>\n<tr>\n<td>20</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>21</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>22</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>23</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>24</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>25</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>26</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>27</td>\n<td>2nd place result_11  776 575</td>\n<td></td>\n</tr>\n<tr>\n<td>28</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td></td>\n</tr>\n<tr>\n<td>29</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>30</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>31</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>32</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>33</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>34</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>35</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>36</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>37</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>38</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>39</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>40</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>41</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>42</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>43</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>44</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>2</td>\n</tr>\n<tr>\n<td>45</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>46</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>47</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>48</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>49</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>50</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>51</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>52</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>53</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>54</td>\n<td>24plase LS NK  784</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>55</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>56</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>57</td>\n<td>24plase LS RG  834</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>58</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>59</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>60</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>61</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>62</td>\n<td>1plase 729 569</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>63</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>64</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>65</td>\n<td>24plase LS NK  784</td>\n<td>2.9</td>\n</tr>\n<tr>\n<td>66</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>67</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>68</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>69</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>70</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>71</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>72</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>73</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>74</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>75</td>\n<td>24plase LS RG  834</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>76</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>77</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>78</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>79</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>80</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>81</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>82</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>83</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>84</td>\n<td>24plase LS NK  784</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>85</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>86</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>87</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>88</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>89</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>90</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>91</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>92</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>93</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>94</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>95</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>96</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>97</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>98</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>99</td>\n<td>2nd place result_15  758 551</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>100</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>101</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>102</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>103</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>104</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>105</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>106</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>107</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>108</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>109</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>110</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>111</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>112</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>113</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>114</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>115</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>116</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>117</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>118</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>119</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>120</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>121</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>122</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>123</td>\n<td>m4_775_LS_T4 775</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>124</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>125</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>126</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>127</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>128</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>129</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>130</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>131</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>132</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>133</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.675</td>\n</tr>\n<tr>\n<td>134</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>135</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>136</td>\n<td>24plase LS NK  784</td>\n<td>2.031</td>\n</tr>\n<tr>\n<td>137</td>\n<td>2nd place result_15  758 551</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>138</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>139</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>140</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>141</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>142</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>143</td>\n<td>3rd place submission 732,2 548</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>144</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>145</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>146</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>147</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.21</td>\n</tr>\n<tr>\n<td>148</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>149</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>150</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td></td>\n</tr>\n<tr>\n<td>151</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>152</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>153</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>154</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>155</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>156</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>157</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td></td>\n</tr>\n<tr>\n<td>158</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td></td>\n</tr>\n<tr>\n<td>159</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>160</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>161</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>162</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>163</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>164</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>165</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.65</td>\n</tr>\n<tr>\n<td>166</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>167</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>168</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>169</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>170</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>171</td>\n<td>pyBoost 760. 586</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>172</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>173</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>174</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>175</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>176</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>177</td>\n<td>m4_775_LS_T4 775</td>\n<td></td>\n</tr>\n<tr>\n<td>178</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>179</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>180</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>181</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>182</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>183</td>\n<td>24plase LS RG  834</td>\n<td>1.5</td>\n</tr>\n<tr>\n<td>184</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>185</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>186</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>187</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>188</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>189</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>190</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td></td>\n</tr>\n<tr>\n<td>191</td>\n<td>24plase LS NK  784</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>192</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>193</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>194</td>\n<td>2nd place result_15  758 551</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>195</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>196</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>197</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>198</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>199</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>200</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>201</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>202</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>203</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td></td>\n</tr>\n<tr>\n<td>204</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>205</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>206</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>207</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.2</td>\n</tr>\n<tr>\n<td>208</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>209</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>210</td>\n<td>24plase LS RG  834</td>\n<td>1.870</td>\n</tr>\n<tr>\n<td>211</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>212</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>213</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>214</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>215</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>216</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>217</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>218</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>219</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>220</td>\n<td>n9_736_3rd_place_psevdo_submission  736</td>\n<td></td>\n</tr>\n<tr>\n<td>221</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>222</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>223</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>224</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>225</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>226</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>227</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>228</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>229</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>230</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>231</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>232</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>233</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>234</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>235</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>236</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>237</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>238</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>239</td>\n<td>JK-PIECE Submit-best 719 592</td>\n<td>1.351</td>\n</tr>\n<tr>\n<td>240</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>241</td>\n<td>m18_790_submission_calibrated_max_d 790</td>\n<td>1.3</td>\n</tr>\n<tr>\n<td>242</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>243</td>\n<td>24plase LS NK  784</td>\n<td></td>\n</tr>\n<tr>\n<td>244</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>245</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>246</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>247</td>\n<td>2nd place result_9 750. 559</td>\n<td></td>\n</tr>\n<tr>\n<td>248</td>\n<td>4th_Place_Ensembling_submission 746. 564</td>\n<td></td>\n</tr>\n<tr>\n<td>249</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>250</td>\n<td>1plase 729 569</td>\n<td></td>\n</tr>\n<tr>\n<td>251</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n<tr>\n<td>252</td>\n<td>pyBoost 760. 586</td>\n<td></td>\n</tr>\n<tr>\n<td>253</td>\n<td>24plase LS RG  834</td>\n<td></td>\n</tr>\n<tr>\n<td>254</td>\n<td>2nd place result_15  758 551</td>\n<td></td>\n</tr>\n<tr>\n<td>255</td>\n<td>3rd place submission 732,2 548</td>\n<td></td>\n</tr>\n</tbody>\n</table>\n<h2>4. Data preprocessing (  solution)</h2>\n<p>The models employed are NN and pyboost.</p>\n<p>Cross validation is not used</p>\n<h2>5. Method (  late submission)</h2>\n<h2>Ensembling <strong>late submissions</strong></h2>\n<p>I evaluate for various compounds the models:</p>\n<table>\n<thead>\n<tr>\n<th>model</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>1th place</td>\n</tr>\n<tr>\n<td>2nd place</td>\n</tr>\n<tr>\n<td>3rd place</td>\n</tr>\n<tr>\n<td>4th place</td>\n</tr>\n<tr>\n<td>24th place NK</td>\n</tr>\n<tr>\n<td>24th place RG</td>\n</tr>\n<tr>\n<td>pyBoost</td>\n</tr>\n</tbody>\n</table>\n<p>Green select indicates that the model has the lowest error among all.</p>\n<table>\n<thead>\n<tr>\n<th>late submissions</th>\n<th>private LB</th>\n<th>public LB</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>1th place+2nd place+3rd place+4th place+24th place NK+24th place RG+pyBoost+50%+mult</td>\n<td><strong>0,597</strong></td>\n<td>0,548</td>\n</tr>\n</tbody>\n</table>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F1dadc65287d17a1469ebbd71e0e95364%2F2023-12-13%20%2023.15.41.png?generation=1702499882391676&amp;alt=media\" alt=\"\"></p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F887cf9a7156a8a35b83850cdf7168c8f%2F2023-12-13%20%2023.16.16.png?generation=1702499901312868&amp;alt=media\" alt=\"\"></p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F8600e331c1e6304df98f7bcaab85a41e%2F2023-12-13%20%2023.16.44.png?generation=1702499920405715&amp;alt=media\" alt=\"\"></p>\n<p>Why is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?</p>\n<h2>6. Final result (  solution,   late submission)</h2>\n<table>\n<thead>\n<tr>\n<th>solution</th>\n<th>private LB</th>\n<th>public LB</th>\n<th>solution or late submission</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>bland</td>\n<td><strong>0,577</strong></td>\n<td>0,53</td>\n<td>late submission</td>\n</tr>\n<tr>\n<td>Pyboost max_depth 10, ntrees 5000, lr 001, colsample 035</td>\n<td>0,763</td>\n<td>0,586</td>\n<td>solution</td>\n</tr>\n<tr>\n<td>Pyboost max_depth 12, ntrees 5000, lr 001, colsample 075,  n_components 49,  T8 b7t17</td>\n<td><strong>0,760</strong></td>\n<td>0,586</td>\n<td>solution</td>\n</tr>\n<tr>\n<td>Pyboost max_depth 12, ntrees 1000, lr 001, colsample 035, n_components 50,   T8 b7t1 bcell7 mie6</td>\n<td>0,761</td>\n<td>0,578</td>\n<td>solution</td>\n</tr>\n<tr>\n<td>quantile(0.7)</td>\n<td>1,003</td>\n<td>0,797</td>\n<td>solution</td>\n</tr>\n</tbody>\n</table>\n<h2>What help (  solution)</h2>\n<p><strong>Drop T8 cell</strong><br>\n<strong>Repeat 17 drug   - 7 time</strong>    +0,001 to Pyboost</p>\n<h2>What help (  late submission)</h2>\n<p>The main score improvement is in <strong>4th Place Magic Postprocessing</strong>  - <br>\n<strong>multiply the compound by a constant</strong> (between 1.4 and 3.09).</p>\n<p>Base score 3rd place submission Private LB 732 , public LB 548</p>\n<table>\n<thead>\n<tr>\n<th>Algorithm</th>\n<th>Private LB  improve</th>\n<th>Private LB</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Base   3rd place submission</td>\n<td></td>\n<td>0.732</td>\n</tr>\n<tr>\n<td>select model from 10 for compound</td>\n<td>+0.075</td>\n<td>0.657</td>\n</tr>\n<tr>\n<td><strong>B cells, ABT-199</strong> =    50% model  <strong>B cells, ABT-199</strong>  + 50% train  <strong>B cells, Alvocidib</strong></td>\n<td>+0.015</td>\n<td></td>\n</tr>\n<tr>\n<td>Add constant (+0.1, -0.1, +3 B cells  BMS-387032)</td>\n<td>+0.008</td>\n<td></td>\n</tr>\n<tr>\n<td>50% model +  50% compound</td>\n<td>+0.008</td>\n<td>0.626</td>\n</tr>\n<tr>\n<td>4th Place Magic Postprocessing of 1 compound: <strong>B cells, BMS-387032</strong>  MULT 3.09</td>\n<td>+0.027</td>\n<td>0.609</td>\n</tr>\n<tr>\n<td>4th Place Magic Postprocessing of 55 compounds MULT 1.2, 1.3, 2</td>\n<td>+0.009</td>\n<td><strong>0.600</strong></td>\n</tr>\n</tbody>\n</table>\n<h2>+0.008 Private LB detail:</h2>\n<p><strong>Compound A</strong> = <strong>Compound A</strong> model * 50% + <strong>Compound B</strong> train * 50%</p>\n<table>\n<thead>\n<tr>\n<th>50% model  submission Cell, Compound</th>\n<th>50% train Cell, Compound</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>27 B cells, CGM-097</td>\n<td>train row 3. T regulatory cells,    Clotrimazole</td>\n</tr>\n<tr>\n<td>65 B cells, Lamivudine</td>\n<td>train row 7. T regulatory cells,    Mometasone Furoate</td>\n</tr>\n<tr>\n<td>147 Myeloid cells, BMS-536924</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>152 Myeloid cells,CC-401</td>\n<td>train row 1. T cells CD4+,  Clotrimazole</td>\n</tr>\n<tr>\n<td>155 Myeloid cells,CGM-097</td>\n<td>train row 0. NK cells,  Clotrimazole</td>\n</tr>\n<tr>\n<td>157 Myeloid cells,Canertinib</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>207    Myeloid cells,PF-03814735</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>214    Myeloid cells,Prednisolone</td>\n<td>train row 6. T cells CD8+,  Mometasone Furoate</td>\n</tr>\n<tr>\n<td>229    Myeloid cells,SLx-2119</td>\n<td>train row 2. T cells CD8+,  Clotrimazole</td>\n</tr>\n<tr>\n<td>241    Myeloid cells,Tacalcitol</td>\n<td>train row 234. Myeloid cells,   Porcn Inhibitor III</td>\n</tr>\n<tr>\n<td>247    Myeloid cells,Tosedostat</td>\n<td>train row 1. T cells CD4+,  Clotrimazole</td>\n</tr>\n<tr>\n<td>248    Myeloid cells,Trametinib</td>\n<td>train row 9. Myeloid cells, Idelalisib</td>\n</tr>\n<tr>\n<td>251    Myeloid cells,Vanoxerine</td>\n<td>train row 1. T cells CD4+,  Clotrimazole</td>\n</tr>\n</tbody>\n</table>\n<h2>7. Sources (  solution)</h2>\n<p><a href=\"https://www.kaggle.com/code/liudacheldieva/blend-for-single-cell-perturbations-dd4fc8\" target=\"_blank\">https://www.kaggle.com/code/liudacheldieva/blend-for-single-cell-perturbations-dd4fc8</a></p>\n<p><strong>0.760</strong> <a href=\"https://www.kaggle.com/code/jeffreylihkust/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/notebook?scriptVersionId=150765833\" target=\"_blank\">https://www.kaggle.com/code/jeffreylihkust/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/notebook?scriptVersionId=150765833</a><br>\n0.761 <a href=\"https://www.kaggle.com/code/emmawilsonev/copy-of-pyboost-secret-grandmaster-s-to-1d68b4?scriptVersionId=150583811\" target=\"_blank\">https://www.kaggle.com/code/emmawilsonev/copy-of-pyboost-secret-grandmaster-s-to-1d68b4?scriptVersionId=150583811</a><br>\n<a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/458750\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/458750</a><br>\n<a href=\"https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/460191\" target=\"_blank\">https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/460191</a></p>\n<h2>8. Sources (  late submission)</h2>\n<p>Private score <strong>0.577</strong>  <a href=\"https://www.kaggle.com/code/liudacheldieva/fork-of-ensembling\" target=\"_blank\">https://www.kaggle.com/code/liudacheldieva/fork-of-ensembling</a> <br>\nsummation.csv 0.577: <a href=\"https://www.kaggle.com/datasets/liudacheldieva/op-0-670\" target=\"_blank\">https://www.kaggle.com/datasets/liudacheldieva/op-0-670</a> </p>\n<h2>9. Used submission.csv (  late submission)</h2>\n<p>import_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')<br>\nimport_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')<br>\nimport_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')<br>\nimport_760pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')<br>\nn1_776_result_11  = pd.read_csv('/kaggle/input/2-pla-submit/result (11).csv', index_col='id')<br>\nn5_760_submission_pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')<br>\nn6_776_submission_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')<br>\nn7_792_submission_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')<br>\nn8_813_submission_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')<br>\nn3_750_result_9 = pd.read_csv('/kaggle/input/2-pla-submit/result (9).csv', index_col='id')<br>\nn4_767_result_8 = pd.read_csv('/kaggle/input/2-pla-submit/result (8).csv', index_col='id')<br>\nn2_758_result_15 = pd.read_csv('/kaggle/input/2-pla-submit/result (15).csv', index_col='id')<br>\nn9_736_3rd_place_psevdo_submission = pd.read_csv('/kaggle/input/3rd-place/3rd place psevdo submission(14).csv', index_col='id')<br>\nn12_746_OP_pub530_myModels_exp1 = pd.read_csv('/kaggle/input/4-pl-sub-late/OP_pub530_myModels_exp1.csv', index_col='id')<br>\nimport_732 = pd.read_csv('/kaggle/input/3rd-place/3rd place submission.csv', index_col='id')<br>\nimport_746 = pd.read_csv('/kaggle/input/4-pl-sub-late/4th_Place_Ensembling_submission.csv', index_col='id')<br>\nn13_729_1_pl = pd.read_csv('/kaggle/input/submit-best/submission.csv', index_col='id')<br>\nn14_719_sub_b = pd.read_csv('/kaggle/input/1st-place-solution/submission.csv', index_col='id')<br>\nimport_670 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729.csv', index_col='id')<br>\nimport_657 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785.csv', index_col='id')<br>\nimport_626 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add.csv', index_col='id')<br>\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')<br>\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')<br>\nm1_792_4th_submission = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')<br>\nm2_784_LS_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_NK.csv', index_col='id')<br>\nm3_834_LS_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_RG.csv', index_col='id')<br>\nm4_775_LS_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T4.csv', index_col='id')<br>\nm5_959_LS_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T8.csv', index_col='id')<br>\nm6_784_SVD_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_NK.csv', index_col='id')<br>\nm7_834_SVD_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_RG.csv', index_col='id')<br>\nm8_775_SVD_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T4.csv', index_col='id')<br>\nm9_959_SVD_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T8.csv', index_col='id')<br>\nm10_773_SVD_b_nk_m_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_nk_m_t4.csv', index_col='id')<br>\nm11_786_SVD_b_t4_m_nk = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_t4_m_nk.csv', index_col='id')<br>\nm12_768_SVD_nk_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4.csv', index_col='id')<br>\nm13_771_SVD_nk_t4_t8_rg = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4_t8_rg.csv', index_col='id')<br>\nm14_746_op_late_submission = pd.read_csv('/kaggle/input/op-late-sub-pub530-mymodels-exp1/submission.csv', index_col='id')<br>\nm15_863_plase30_submission = pd.read_csv('/kaggle/input/30-op2scp-transformer-deberta-v3-small-demo/submission.csv', index_col='id')<br>\nm17_769_submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50 = pd.read_csv('/kaggle/input/pyboost-u900/submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')<br>\nm18_790_submission_calibrated_max_d = pd.read_csv('/kaggle/input/pyboost-u900/submission_calibrated_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')<br>\nM19_785_13th_place_mlp = pd.read_csv('/kaggle/input/13th-place-solution-mlp/submission_example_MLP.csv', index_col='id')<br>\nimport_755_10pl = pd.read_csv('/kaggle/input/10th-place/sub_submission.csv', index_col='id')<br>\nimport_748_26pl = pd.read_csv('/kaggle/input/op-26-place/submission.csv', index_col='id')<br>\nimport_577 = pd.read_csv('/kaggle/input/op-0-670/submi_mu_0.577.csv', index_col='id')</p>\n<p>Thank you for taking the time to read. Feel free to ask any questions you may have.</p>",
      "rawMarkdown": "## 1. Introduction\nI want to thank Kaggle and everyone involved for hosting such an intriguing competition.\n\nTwo topics are what I describe.\n1.1.   late submission\n1.2.   solution \n\nIt seems that one of the models is better for one compound but weak for other compound.\n\nThe question remains: how to choose the right drug model for compound?\n\nConsidering that cells of various types have different models.\n\nCould you please clarify why a model is good for a compound?\nWhy is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?\n\nInitially, it is necessary to cover a larger portion of the error with models. \nWith an error of 0.661 and without knowing which compounds have large errors, it is difficult to say whether the model is suitable for the compound.\n\n## 2. Context\n\nBusiness context: https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/overview\nData context: https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/data\n\n## 3. Overview of the approach (  late submission)\n\nI like to discover which model is better for compounds.\nThe result is better if you select one model for each drug and cell type.\nIt's possible that there's a chemical explanation for model selection.\n\nWe lack the skill to select from all the models available.\nAll models have been added.\n| bland | Private score |\n| --- | --- |\n| Weight sum  of 10 models   | 0.718 |\n| Select one model for each compound | 0.657 |\n| Multiply the compound by a constant (between 1.4 and 3.09)| 0.576 |\n\n\n### Late submission 0.577:\n|\tSubmition id\t|\tModel\t|\tMultiplecator is Post Processing \t|\n| --- | --- |--- |\n|\t0\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t1\t|\t2nd place result_11  776 575 *50% +\t+ 50% train  **B cells, Alvocidib** |\t\t|\n|\t2\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t3\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t4\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t5\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t6\t|\tm4_775_LS_T4 775\t|\t 1.2\t|\n|\t7\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t8\t|\t24plase LS RG  834\t|\t\t|\n|\t9\t|\t24plase LS NK  784\t|\t 1.2\t|\n|\t10\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.2\t|\n|\t11\t|\tm18_790_submission_calibrated_max_d 790\t|\t\t|\n|\t12\t|\tm18_790_submission_calibrated_max_d 790\t|\t\t|\n|\t13\t|\t24plase LS NK  784\t|\t\t|\n|\t14\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t15\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.2\t|\n|\t16\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t17\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t18\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t19\t|\t1plase 729 569\t|\t 3.09\t|\n|\t20\t|\t1plase 729 569\t|\t\t|\n|\t21\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t22\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t23\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.3\t|\n|\t24\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t25\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t26\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t27\t|\t2nd place result_11  776 575\t|\t\t|\n|\t28\t|\tJK-PIECE Submit-best 719 592\t|\t\t|\n|\t29\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t30\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.2\t|\n|\t31\t|\t1plase 729 569\t|\t\t|\n|\t32\t|\tpyBoost 760. 586\t|\t\t|\n|\t33\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t34\t|\t24plase LS NK  784\t|\t\t|\n|\t35\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t36\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t37\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t38\t|\tpyBoost 760. 586\t|\t\t|\n|\t39\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t40\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t41\t|\tpyBoost 760. 586\t|\t\t|\n|\t42\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.2\t|\n|\t43\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t44\t|\tm18_790_submission_calibrated_max_d 790\t|\t2\t|\n|\t45\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t46\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t47\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t48\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t49\t|\t24plase LS NK  784\t|\t\t|\n|\t50\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.2\t|\n|\t51\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t52\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t53\t|\t24plase LS RG  834\t|\t\t|\n|\t54\t|\t24plase LS NK  784\t|\t 1.3\t|\n|\t55\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t56\t|\t1plase 729 569\t|\t\t|\n|\t57\t|\t24plase LS RG  834\t|\t 1.2\t|\n|\t58\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t59\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t60\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t61\t|\t24plase LS NK  784\t|\t\t|\n|\t62\t|\t1plase 729 569\t|\t 1.2\t|\n|\t63\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t64\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.3\t|\n|\t65\t|\t24plase LS NK  784\t|\t 2.9\t|\n|\t66\t|\t24plase LS NK  784\t|\t\t|\n|\t67\t|\tm4_775_LS_T4 775\t|\t 1.2\t|\n|\t68\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t69\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t70\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t71\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t72\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t73\t|\t24plase LS NK  784\t|\t\t|\n|\t74\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t75\t|\t24plase LS RG  834\t|\t 1.3\t|\n|\t76\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.3\t|\n|\t77\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t78\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t79\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t80\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.2\t|\n|\t81\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t82\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t83\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t84\t|\t24plase LS NK  784\t|\t 1.3\t|\n|\t85\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t86\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t87\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.3\t|\n|\t88\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t89\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t90\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t91\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t92\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t93\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t94\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.3\t|\n|\t95\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t96\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t97\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t98\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t99\t|\t2nd place result_15  758 551\t|\t 1.2\t|\n|\t100\t|\t24plase LS NK  784\t|\t\t|\n|\t101\t|\t2nd place result_15  758 551\t|\t\t|\n|\t102\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t103\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t104\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t105\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t106\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t107\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t108\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t109\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t110\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t111\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t112\t|\t1plase 729 569\t|\t\t|\n|\t113\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t114\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t115\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t116\t|\tm4_775_LS_T4 775\t|\t 1.2\t|\n|\t117\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t118\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t119\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t120\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t121\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t122\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t123\t|\tm4_775_LS_T4 775\t|\t 1.3\t|\n|\t124\t|\t24plase LS RG  834\t|\t\t|\n|\t125\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t126\t|\t1plase 729 569\t|\t\t|\n|\t127\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t128\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t129\t|\t24plase LS NK  784\t|\t\t|\n|\t130\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t131\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t132\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t133\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.675\t|\n|\t134\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t135\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t136\t|\t24plase LS NK  784\t|\t 2.031\t|\n|\t137\t|\t2nd place result_15  758 551\t|\t 1.2\t|\n|\t138\t|\t2nd place result_15  758 551\t|\t\t|\n|\t139\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t140\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t141\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t142\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t143\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t144\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t145\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t146\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t147\t|\tJK-PIECE Submit-best 719 592\t|\t 1.21\t|\n|\t148\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t149\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t150\t|\tJK-PIECE Submit-best 719 592\t|\t\t|\n|\t151\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t152\t|\t1plase 729 569\t|\t\t|\n|\t153\t|\t1plase 729 569\t|\t\t|\n|\t154\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t155\t|\t2nd place result_15  758 551\t|\t\t|\n|\t156\t|\tpyBoost 760. 586\t|\t\t|\n|\t157\t|\tm18_790_submission_calibrated_max_d 790\t|\t\t|\n|\t158\t|\tJK-PIECE Submit-best 719 592\t|\t\t|\n|\t159\t|\tpyBoost 760. 586\t|\t\t|\n|\t160\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t161\t|\tpyBoost 760. 586\t|\t\t|\n|\t162\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t163\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t164\t|\tpyBoost 760. 586\t|\t\t|\n|\t165\t|\tJK-PIECE Submit-best 719 592\t|\t 1.65\t|\n|\t166\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t167\t|\t1plase 729 569\t|\t\t|\n|\t168\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t169\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.2\t|\n|\t170\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t171\t|\tpyBoost 760. 586\t|\t 1.2\t|\n|\t172\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t173\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t174\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t175\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t176\t|\tpyBoost 760. 586\t|\t\t|\n|\t177\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t178\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t179\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t180\t|\tpyBoost 760. 586\t|\t\t|\n|\t181\t|\t2nd place result_15  758 551\t|\t\t|\n|\t182\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t183\t|\t24plase LS RG  834\t|\t 1.5\t|\n|\t184\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t185\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t186\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t187\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t188\t|\tpyBoost 760. 586\t|\t\t|\n|\t189\t|\t24plase LS NK  784\t|\t\t|\n|\t190\t|\t4th_Place_Ensembling_submission 746. 564\t|\t\t|\n|\t191\t|\t24plase LS NK  784\t|\t 1.3\t|\n|\t192\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t193\t|\t2nd place result_15  758 551\t|\t\t|\n|\t194\t|\t2nd place result_15  758 551\t|\t 1.2\t|\n|\t195\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t196\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t197\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t198\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t199\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t200\t|\t24plase LS NK  784\t|\t\t|\n|\t201\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t202\t|\t1plase 729 569\t|\t\t|\n|\t203\t|\t4th_Place_Ensembling_submission 746. 564\t|\t\t|\n|\t204\t|\t2nd place result_15  758 551\t|\t\t|\n|\t205\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t206\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t207\t|\tJK-PIECE Submit-best 719 592\t|\t 1.2\t|\n|\t208\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t209\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t210\t|\t24plase LS RG  834\t|\t 1.870\t|\n|\t211\t|\tpyBoost 760. 586\t|\t\t|\n|\t212\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t213\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t214\t|\t1plase 729 569\t|\t\t|\n|\t215\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t216\t|\tpyBoost 760. 586\t|\t\t|\n|\t217\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t218\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t219\t|\tpyBoost 760. 586\t|\t\t|\n|\t220\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t221\t|\t2nd place result_15  758 551\t|\t\t|\n|\t222\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t223\t|\t24plase LS NK  784\t|\t\t|\n|\t224\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t225\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t226\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t227\t|\t24plase LS NK  784\t|\t\t|\n|\t228\t|\t2nd place result_15  758 551\t|\t\t|\n|\t229\t|\t24plase LS NK  784\t|\t\t|\n|\t230\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t231\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t232\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t233\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t234\t|\tpyBoost 760. 586\t|\t\t|\n|\t235\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t236\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t237\t|\tpyBoost 760. 586\t|\t\t|\n|\t238\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t239\t|\tJK-PIECE Submit-best 719 592\t|\t 1.351\t|\n|\t240\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t241\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.3\t|\n|\t242\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t243\t|\t24plase LS NK  784\t|\t\t|\n|\t244\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t245\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t246\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t247\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t248\t|\t4th_Place_Ensembling_submission 746. 564\t|\t\t|\n|\t249\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t250\t|\t1plase 729 569\t|\t\t|\n|\t251\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t252\t|\tpyBoost 760. 586\t|\t\t|\n|\t253\t|\t24plase LS RG  834\t|\t\t|\n|\t254\t|\t2nd place result_15  758 551\t|\t\t|\n|\t255\t|\t3rd place submission 732,2 548\t|\t\t|\n\n## 4. Data preprocessing (  solution)\n\nThe models employed are NN and pyboost.\n\nCross validation is not used\n\n## 5. Method (  late submission)\n \n## Ensembling **late submissions**\n\nI evaluate for various compounds the models:\n|model|\n| --- |\n|1th place|\n|2nd place|\n|3rd place|\n|4th place|\n|24th place NK|\n|24th place RG|\n|pyBoost|\n\nGreen select indicates that the model has the lowest error among all.\n\n| late submissions | private LB   | public LB |\n| --- | --- | --- |\n| 1th place+2nd place+3rd place+4th place+24th place NK+24th place RG+pyBoost+50%+mult| **0,597** |0,548| \n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F1dadc65287d17a1469ebbd71e0e95364%2F2023-12-13%20%2023.15.41.png?generation=1702499882391676&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F887cf9a7156a8a35b83850cdf7168c8f%2F2023-12-13%20%2023.16.16.png?generation=1702499901312868&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F8600e331c1e6304df98f7bcaab85a41e%2F2023-12-13%20%2023.16.44.png?generation=1702499920405715&alt=media)\n\nWhy is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?\n\n\n## 6. Final result (  solution,   late submission)\n| solution | private LB   | public LB | solution or late submission |\n| --- | --- | --- | --- |\n|bland   | **0,577**|\t  0,53|late submission|\n|Pyboost max_depth 10, ntrees 5000, lr 001, colsample 035  |   0,763\t|  0,586|solution|\n|Pyboost max_depth 12, ntrees 5000, lr 001, colsample 075,  n_components 49,  T8 b7t17  |   **0,760**\t|  0,586|solution|\n|Pyboost max_depth 12, ntrees 1000, lr 001, colsample 035, n_components 50,   T8 b7t1 bcell7 mie6  |   0,761\t|  0,578|solution|\n|quantile(0.7)  |       1,003\t |  0,797|solution|\n\n\n## What help (  solution)\n**Drop T8 cell**\n**Repeat 17 drug   - 7 time**    +0,001 to Pyboost\n\n\n \n##   What help (  late submission)\nThe main score improvement is in **4th Place Magic Postprocessing**  - \n**multiply the compound by a constant** (between 1.4 and 3.09).\n \nBase score 3rd place submission Private LB 732 , public LB 548\n\n| Algorithm | Private LB  improve  |Private LB    |\n| --- | --- |\n|  Base   3rd place submission |   | 0.732 |\n| select model from 10 for compound | +0.075 | 0.657 |\n| **B cells, ABT-199** =    50% model  **B cells, ABT-199**  + 50% train  **B cells, Alvocidib**  |   +0.015 | |\n| Add constant (+0.1, -0.1, +3 B cells  BMS-387032)  | +0.008 | |\n| 50% model +  50% compound | +0.008 |  0.626   |\n| 4th Place Magic Postprocessing of 1 compound: **B cells, BMS-387032**  MULT 3.09 | +0.027 | 0.609  |\n| 4th Place Magic Postprocessing of 55 compounds MULT 1.2, 1.3, 2 | +0.009 | **0.600** |\n\n## +0.008 Private LB detail:\n**Compound A** = **Compound A** model * 50% + **Compound B** train * 50%\n|50% model  submission Cell, Compound  |  50% train Cell, Compound |\n| --- | --- |\n|27 B cells, CGM-097| \t\t\t\t\ttrain row 3. T regulatory cells,\tClotrimazole|\n|65 B cells, Lamivudine|\t \t\t\t\ttrain row 7. T regulatory cells,\tMometasone Furoate|\n|147 Myeloid cells, BMS-536924\t|\t\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|152 Myeloid cells,CC-401|\t\t\t\t\ttrain row 1. T cells CD4+,\tClotrimazole|\n|155 Myeloid cells,CGM-097\t|\t \t\t\ttrain row 0. NK cells,\tClotrimazole|\n|157 Myeloid cells,Canertinib|\t \t\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|207\tMyeloid cells,PF-03814735\t|\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|214\tMyeloid cells,Prednisolone\t|\t\t\ttrain row 6. T cells CD8+,\tMometasone Furoate|\n|229\tMyeloid cells,SLx-2119\t|\t\t\ttrain row 2. T cells CD8+,\tClotrimazole|\n|241\tMyeloid cells,Tacalcitol\t\t|\t\ttrain row 234. Myeloid cells,\tPorcn Inhibitor III|\n|247\tMyeloid cells,Tosedostat|\t\t\t\ttrain row 1. T cells CD4+,\tClotrimazole|\n|248\tMyeloid cells,Trametinib|\t\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|251\tMyeloid cells,Vanoxerine\t|\t\t\ttrain row 1. T cells CD4+,\tClotrimazole|\n\n \n\n## 7. Sources (  solution)\n\nhttps://www.kaggle.com/code/liudacheldieva/blend-for-single-cell-perturbations-dd4fc8\n\n**0.760** https://www.kaggle.com/code/jeffreylihkust/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/notebook?scriptVersionId=150765833\n0.761 https://www.kaggle.com/code/emmawilsonev/copy-of-pyboost-secret-grandmaster-s-to-1d68b4?scriptVersionId=150583811\nhttps://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/458750\nhttps://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/460191\n\n\n## 8. Sources (  late submission)\n\nPrivate score **0.577**  https://www.kaggle.com/code/liudacheldieva/fork-of-ensembling \nsummation.csv 0.577: https://www.kaggle.com/datasets/liudacheldieva/op-0-670 \n\n## 9. Used submission.csv (  late submission)\nimport_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')\nimport_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')\nimport_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')\nimport_760pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')\nn1_776_result_11  = pd.read_csv('/kaggle/input/2-pla-submit/result (11).csv', index_col='id')\nn5_760_submission_pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')\nn6_776_submission_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')\nn7_792_submission_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')\nn8_813_submission_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')\nn3_750_result_9 = pd.read_csv('/kaggle/input/2-pla-submit/result (9).csv', index_col='id')\nn4_767_result_8 = pd.read_csv('/kaggle/input/2-pla-submit/result (8).csv', index_col='id')\nn2_758_result_15 = pd.read_csv('/kaggle/input/2-pla-submit/result (15).csv', index_col='id')\nn9_736_3rd_place_psevdo_submission = pd.read_csv('/kaggle/input/3rd-place/3rd place psevdo submission(14).csv', index_col='id')\nn12_746_OP_pub530_myModels_exp1 = pd.read_csv('/kaggle/input/4-pl-sub-late/OP_pub530_myModels_exp1.csv', index_col='id')\nimport_732 = pd.read_csv('/kaggle/input/3rd-place/3rd place submission.csv', index_col='id')\nimport_746 = pd.read_csv('/kaggle/input/4-pl-sub-late/4th_Place_Ensembling_submission.csv', index_col='id')\nn13_729_1_pl = pd.read_csv('/kaggle/input/submit-best/submission.csv', index_col='id')\nn14_719_sub_b = pd.read_csv('/kaggle/input/1st-place-solution/submission.csv', index_col='id')\nimport_670 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729.csv', index_col='id')\nimport_657 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785.csv', index_col='id')\nimport_626 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add.csv', index_col='id')\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')\nm1_792_4th_submission = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')\nm2_784_LS_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_NK.csv', index_col='id')\nm3_834_LS_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_RG.csv', index_col='id')\nm4_775_LS_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T4.csv', index_col='id')\nm5_959_LS_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T8.csv', index_col='id')\nm6_784_SVD_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_NK.csv', index_col='id')\nm7_834_SVD_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_RG.csv', index_col='id')\nm8_775_SVD_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T4.csv', index_col='id')\nm9_959_SVD_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T8.csv', index_col='id')\nm10_773_SVD_b_nk_m_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_nk_m_t4.csv', index_col='id')\nm11_786_SVD_b_t4_m_nk = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_t4_m_nk.csv', index_col='id')\nm12_768_SVD_nk_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4.csv', index_col='id')\nm13_771_SVD_nk_t4_t8_rg = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4_t8_rg.csv', index_col='id')\nm14_746_op_late_submission = pd.read_csv('/kaggle/input/op-late-sub-pub530-mymodels-exp1/submission.csv', index_col='id')\nm15_863_plase30_submission = pd.read_csv('/kaggle/input/30-op2scp-transformer-deberta-v3-small-demo/submission.csv', index_col='id')\nm17_769_submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50 = pd.read_csv('/kaggle/input/pyboost-u900/submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')\nm18_790_submission_calibrated_max_d = pd.read_csv('/kaggle/input/pyboost-u900/submission_calibrated_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')\nM19_785_13th_place_mlp = pd.read_csv('/kaggle/input/13th-place-solution-mlp/submission_example_MLP.csv', index_col='id')\nimport_755_10pl = pd.read_csv('/kaggle/input/10th-place/sub_submission.csv', index_col='id')\nimport_748_26pl = pd.read_csv('/kaggle/input/op-26-place/submission.csv', index_col='id')\nimport_577 = pd.read_csv('/kaggle/input/op-0-670/submi_mu_0.577.csv', index_col='id')\n\nThank you for taking the time to read. Feel free to ask any questions you may have.",
      "votes": null
    },
    {
      "id": "2558915",
      "postDate": "12/12/2023 12:48:35",
      "content": "<p>Great work!! Congratulations 🎉 </p>",
      "rawMarkdown": "Great work!! Congratulations 🎉",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2558915,
      "author_name": "anthonytherrien",
      "author_url": "",
      "post_date": "12/12/2023 12:48:35",
      "content": "<p>Great work!! Congratulations 🎉 </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2558901": "## 1. Introduction\nI want to thank Kaggle and everyone involved for hosting such an intriguing competition.\n\nTwo topics are what I describe.\n1.1.   late submission\n1.2.   solution \n\nIt seems that one of the models is better for one compound but weak for other compound.\n\nThe question remains: how to choose the right drug model for compound?\n\nConsidering that cells of various types have different models.\n\nCould you please clarify why a model is good for a compound?\nWhy is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?\n\nInitially, it is necessary to cover a larger portion of the error with models. \nWith an error of 0.661 and without knowing which compounds have large errors, it is difficult to say whether the model is suitable for the compound.\n\n## 2. Context\n\nBusiness context: https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/overview\nData context: https://www.kaggle.com/competitions/open-problems-single-cell-perturbations/data\n\n## 3. Overview of the approach (  late submission)\n\nI like to discover which model is better for compounds.\nThe result is better if you select one model for each drug and cell type.\nIt's possible that there's a chemical explanation for model selection.\n\nWe lack the skill to select from all the models available.\nAll models have been added.\n| bland | Private score |\n| --- | --- |\n| Weight sum  of 10 models   | 0.718 |\n| Select one model for each compound | 0.657 |\n| Multiply the compound by a constant (between 1.4 and 3.09)| 0.576 |\n\n\n### Late submission 0.577:\n|\tSubmition id\t|\tModel\t|\tMultiplecator is Post Processing \t|\n| --- | --- |--- |\n|\t0\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t1\t|\t2nd place result_11  776 575 *50% +\t+ 50% train  **B cells, Alvocidib** |\t\t|\n|\t2\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t3\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t4\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t5\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t6\t|\tm4_775_LS_T4 775\t|\t 1.2\t|\n|\t7\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t8\t|\t24plase LS RG  834\t|\t\t|\n|\t9\t|\t24plase LS NK  784\t|\t 1.2\t|\n|\t10\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.2\t|\n|\t11\t|\tm18_790_submission_calibrated_max_d 790\t|\t\t|\n|\t12\t|\tm18_790_submission_calibrated_max_d 790\t|\t\t|\n|\t13\t|\t24plase LS NK  784\t|\t\t|\n|\t14\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t15\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.2\t|\n|\t16\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t17\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t18\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t19\t|\t1plase 729 569\t|\t 3.09\t|\n|\t20\t|\t1plase 729 569\t|\t\t|\n|\t21\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t22\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t23\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.3\t|\n|\t24\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t25\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t26\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t27\t|\t2nd place result_11  776 575\t|\t\t|\n|\t28\t|\tJK-PIECE Submit-best 719 592\t|\t\t|\n|\t29\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t30\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.2\t|\n|\t31\t|\t1plase 729 569\t|\t\t|\n|\t32\t|\tpyBoost 760. 586\t|\t\t|\n|\t33\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t34\t|\t24plase LS NK  784\t|\t\t|\n|\t35\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t36\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t37\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t38\t|\tpyBoost 760. 586\t|\t\t|\n|\t39\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t40\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t41\t|\tpyBoost 760. 586\t|\t\t|\n|\t42\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.2\t|\n|\t43\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t44\t|\tm18_790_submission_calibrated_max_d 790\t|\t2\t|\n|\t45\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t46\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t47\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t48\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t49\t|\t24plase LS NK  784\t|\t\t|\n|\t50\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.2\t|\n|\t51\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t52\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t53\t|\t24plase LS RG  834\t|\t\t|\n|\t54\t|\t24plase LS NK  784\t|\t 1.3\t|\n|\t55\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t56\t|\t1plase 729 569\t|\t\t|\n|\t57\t|\t24plase LS RG  834\t|\t 1.2\t|\n|\t58\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t59\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t60\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t61\t|\t24plase LS NK  784\t|\t\t|\n|\t62\t|\t1plase 729 569\t|\t 1.2\t|\n|\t63\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t64\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.3\t|\n|\t65\t|\t24plase LS NK  784\t|\t 2.9\t|\n|\t66\t|\t24plase LS NK  784\t|\t\t|\n|\t67\t|\tm4_775_LS_T4 775\t|\t 1.2\t|\n|\t68\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t69\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t70\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t71\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t72\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t73\t|\t24plase LS NK  784\t|\t\t|\n|\t74\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t75\t|\t24plase LS RG  834\t|\t 1.3\t|\n|\t76\t|\t4th_Place_Ensembling_submission 746. 564\t|\t 1.3\t|\n|\t77\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t78\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t79\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t80\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.2\t|\n|\t81\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t82\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t83\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t84\t|\t24plase LS NK  784\t|\t 1.3\t|\n|\t85\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t86\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t87\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.3\t|\n|\t88\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t89\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t90\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t91\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t92\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t93\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t94\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t 1.3\t|\n|\t95\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t96\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t97\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t98\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t99\t|\t2nd place result_15  758 551\t|\t 1.2\t|\n|\t100\t|\t24plase LS NK  784\t|\t\t|\n|\t101\t|\t2nd place result_15  758 551\t|\t\t|\n|\t102\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t103\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t104\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t105\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t106\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t107\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t108\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t109\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t110\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t111\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t112\t|\t1plase 729 569\t|\t\t|\n|\t113\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t114\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t115\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t116\t|\tm4_775_LS_T4 775\t|\t 1.2\t|\n|\t117\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t118\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t119\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t120\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t121\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t122\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t123\t|\tm4_775_LS_T4 775\t|\t 1.3\t|\n|\t124\t|\t24plase LS RG  834\t|\t\t|\n|\t125\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t126\t|\t1plase 729 569\t|\t\t|\n|\t127\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t128\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t129\t|\t24plase LS NK  784\t|\t\t|\n|\t130\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t131\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t132\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t133\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.675\t|\n|\t134\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t135\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t136\t|\t24plase LS NK  784\t|\t 2.031\t|\n|\t137\t|\t2nd place result_15  758 551\t|\t 1.2\t|\n|\t138\t|\t2nd place result_15  758 551\t|\t\t|\n|\t139\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t140\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t141\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t142\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t143\t|\t3rd place submission 732,2 548\t|\t 1.2\t|\n|\t144\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t145\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t146\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t147\t|\tJK-PIECE Submit-best 719 592\t|\t 1.21\t|\n|\t148\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t149\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t150\t|\tJK-PIECE Submit-best 719 592\t|\t\t|\n|\t151\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t152\t|\t1plase 729 569\t|\t\t|\n|\t153\t|\t1plase 729 569\t|\t\t|\n|\t154\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t155\t|\t2nd place result_15  758 551\t|\t\t|\n|\t156\t|\tpyBoost 760. 586\t|\t\t|\n|\t157\t|\tm18_790_submission_calibrated_max_d 790\t|\t\t|\n|\t158\t|\tJK-PIECE Submit-best 719 592\t|\t\t|\n|\t159\t|\tpyBoost 760. 586\t|\t\t|\n|\t160\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t161\t|\tpyBoost 760. 586\t|\t\t|\n|\t162\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t163\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t164\t|\tpyBoost 760. 586\t|\t\t|\n|\t165\t|\tJK-PIECE Submit-best 719 592\t|\t 1.65\t|\n|\t166\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t167\t|\t1plase 729 569\t|\t\t|\n|\t168\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t169\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.2\t|\n|\t170\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t171\t|\tpyBoost 760. 586\t|\t 1.2\t|\n|\t172\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t173\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t174\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t175\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t176\t|\tpyBoost 760. 586\t|\t\t|\n|\t177\t|\tm4_775_LS_T4 775\t|\t\t|\n|\t178\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t179\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t180\t|\tpyBoost 760. 586\t|\t\t|\n|\t181\t|\t2nd place result_15  758 551\t|\t\t|\n|\t182\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t183\t|\t24plase LS RG  834\t|\t 1.5\t|\n|\t184\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t185\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t186\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t187\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t188\t|\tpyBoost 760. 586\t|\t\t|\n|\t189\t|\t24plase LS NK  784\t|\t\t|\n|\t190\t|\t4th_Place_Ensembling_submission 746. 564\t|\t\t|\n|\t191\t|\t24plase LS NK  784\t|\t 1.3\t|\n|\t192\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t193\t|\t2nd place result_15  758 551\t|\t\t|\n|\t194\t|\t2nd place result_15  758 551\t|\t 1.2\t|\n|\t195\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t196\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t197\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t198\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t199\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t200\t|\t24plase LS NK  784\t|\t\t|\n|\t201\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t202\t|\t1plase 729 569\t|\t\t|\n|\t203\t|\t4th_Place_Ensembling_submission 746. 564\t|\t\t|\n|\t204\t|\t2nd place result_15  758 551\t|\t\t|\n|\t205\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t206\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t207\t|\tJK-PIECE Submit-best 719 592\t|\t 1.2\t|\n|\t208\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t209\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t210\t|\t24plase LS RG  834\t|\t 1.870\t|\n|\t211\t|\tpyBoost 760. 586\t|\t\t|\n|\t212\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t213\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t214\t|\t1plase 729 569\t|\t\t|\n|\t215\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t216\t|\tpyBoost 760. 586\t|\t\t|\n|\t217\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t218\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t219\t|\tpyBoost 760. 586\t|\t\t|\n|\t220\t|\tn9_736_3rd_place_psevdo_submission  736\t|\t\t|\n|\t221\t|\t2nd place result_15  758 551\t|\t\t|\n|\t222\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t223\t|\t24plase LS NK  784\t|\t\t|\n|\t224\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t225\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t226\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t227\t|\t24plase LS NK  784\t|\t\t|\n|\t228\t|\t2nd place result_15  758 551\t|\t\t|\n|\t229\t|\t24plase LS NK  784\t|\t\t|\n|\t230\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t231\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t232\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t233\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t234\t|\tpyBoost 760. 586\t|\t\t|\n|\t235\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t236\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t237\t|\tpyBoost 760. 586\t|\t\t|\n|\t238\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t239\t|\tJK-PIECE Submit-best 719 592\t|\t 1.351\t|\n|\t240\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t241\t|\tm18_790_submission_calibrated_max_d 790\t|\t 1.3\t|\n|\t242\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t243\t|\t24plase LS NK  784\t|\t\t|\n|\t244\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t245\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t246\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t247\t|\t2nd place result_9 750. 559\t|\t\t|\n|\t248\t|\t4th_Place_Ensembling_submission 746. 564\t|\t\t|\n|\t249\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t250\t|\t1plase 729 569\t|\t\t|\n|\t251\t|\t3rd place submission 732,2 548\t|\t\t|\n|\t252\t|\tpyBoost 760. 586\t|\t\t|\n|\t253\t|\t24plase LS RG  834\t|\t\t|\n|\t254\t|\t2nd place result_15  758 551\t|\t\t|\n|\t255\t|\t3rd place submission 732,2 548\t|\t\t|\n\n## 4. Data preprocessing (  solution)\n\nThe models employed are NN and pyboost.\n\nCross validation is not used\n\n## 5. Method (  late submission)\n \n## Ensembling **late submissions**\n\nI evaluate for various compounds the models:\n|model|\n| --- |\n|1th place|\n|2nd place|\n|3rd place|\n|4th place|\n|24th place NK|\n|24th place RG|\n|pyBoost|\n\nGreen select indicates that the model has the lowest error among all.\n\n| late submissions | private LB   | public LB |\n| --- | --- | --- |\n| 1th place+2nd place+3rd place+4th place+24th place NK+24th place RG+pyBoost+50%+mult| **0,597** |0,548| \n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F1dadc65287d17a1469ebbd71e0e95364%2F2023-12-13%20%2023.15.41.png?generation=1702499882391676&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F887cf9a7156a8a35b83850cdf7168c8f%2F2023-12-13%20%2023.16.16.png?generation=1702499901312868&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4172517%2F8600e331c1e6304df98f7bcaab85a41e%2F2023-12-13%20%2023.16.44.png?generation=1702499920405715&alt=media)\n\nWhy is Vorinostat's '24 place model LS RG' the best model for Myeloid cells, with a private score improvement of +0.003?\n\n\n## 6. Final result (  solution,   late submission)\n| solution | private LB   | public LB | solution or late submission |\n| --- | --- | --- | --- |\n|bland   | **0,577**|\t  0,53|late submission|\n|Pyboost max_depth 10, ntrees 5000, lr 001, colsample 035  |   0,763\t|  0,586|solution|\n|Pyboost max_depth 12, ntrees 5000, lr 001, colsample 075,  n_components 49,  T8 b7t17  |   **0,760**\t|  0,586|solution|\n|Pyboost max_depth 12, ntrees 1000, lr 001, colsample 035, n_components 50,   T8 b7t1 bcell7 mie6  |   0,761\t|  0,578|solution|\n|quantile(0.7)  |       1,003\t |  0,797|solution|\n\n\n## What help (  solution)\n**Drop T8 cell**\n**Repeat 17 drug   - 7 time**    +0,001 to Pyboost\n\n\n \n##   What help (  late submission)\nThe main score improvement is in **4th Place Magic Postprocessing**  - \n**multiply the compound by a constant** (between 1.4 and 3.09).\n \nBase score 3rd place submission Private LB 732 , public LB 548\n\n| Algorithm | Private LB  improve  |Private LB    |\n| --- | --- |\n|  Base   3rd place submission |   | 0.732 |\n| select model from 10 for compound | +0.075 | 0.657 |\n| **B cells, ABT-199** =    50% model  **B cells, ABT-199**  + 50% train  **B cells, Alvocidib**  |   +0.015 | |\n| Add constant (+0.1, -0.1, +3 B cells  BMS-387032)  | +0.008 | |\n| 50% model +  50% compound | +0.008 |  0.626   |\n| 4th Place Magic Postprocessing of 1 compound: **B cells, BMS-387032**  MULT 3.09 | +0.027 | 0.609  |\n| 4th Place Magic Postprocessing of 55 compounds MULT 1.2, 1.3, 2 | +0.009 | **0.600** |\n\n## +0.008 Private LB detail:\n**Compound A** = **Compound A** model * 50% + **Compound B** train * 50%\n|50% model  submission Cell, Compound  |  50% train Cell, Compound |\n| --- | --- |\n|27 B cells, CGM-097| \t\t\t\t\ttrain row 3. T regulatory cells,\tClotrimazole|\n|65 B cells, Lamivudine|\t \t\t\t\ttrain row 7. T regulatory cells,\tMometasone Furoate|\n|147 Myeloid cells, BMS-536924\t|\t\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|152 Myeloid cells,CC-401|\t\t\t\t\ttrain row 1. T cells CD4+,\tClotrimazole|\n|155 Myeloid cells,CGM-097\t|\t \t\t\ttrain row 0. NK cells,\tClotrimazole|\n|157 Myeloid cells,Canertinib|\t \t\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|207\tMyeloid cells,PF-03814735\t|\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|214\tMyeloid cells,Prednisolone\t|\t\t\ttrain row 6. T cells CD8+,\tMometasone Furoate|\n|229\tMyeloid cells,SLx-2119\t|\t\t\ttrain row 2. T cells CD8+,\tClotrimazole|\n|241\tMyeloid cells,Tacalcitol\t\t|\t\ttrain row 234. Myeloid cells,\tPorcn Inhibitor III|\n|247\tMyeloid cells,Tosedostat|\t\t\t\ttrain row 1. T cells CD4+,\tClotrimazole|\n|248\tMyeloid cells,Trametinib|\t\t\t\ttrain row 9. Myeloid cells,\tIdelalisib|\n|251\tMyeloid cells,Vanoxerine\t|\t\t\ttrain row 1. T cells CD4+,\tClotrimazole|\n\n \n\n## 7. Sources (  solution)\n\nhttps://www.kaggle.com/code/liudacheldieva/blend-for-single-cell-perturbations-dd4fc8\n\n**0.760** https://www.kaggle.com/code/jeffreylihkust/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/notebook?scriptVersionId=150765833\n0.761 https://www.kaggle.com/code/emmawilsonev/copy-of-pyboost-secret-grandmaster-s-to-1d68b4?scriptVersionId=150583811\nhttps://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/458750\nhttps://www.kaggle.com/competitions/open-problems-single-cell-perturbations/discussion/460191\n\n\n## 8. Sources (  late submission)\n\nPrivate score **0.577**  https://www.kaggle.com/code/liudacheldieva/fork-of-ensembling \nsummation.csv 0.577: https://www.kaggle.com/datasets/liudacheldieva/op-0-670 \n\n## 9. Used submission.csv (  late submission)\nimport_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')\nimport_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')\nimport_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')\nimport_760pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')\nn1_776_result_11  = pd.read_csv('/kaggle/input/2-pla-submit/result (11).csv', index_col='id')\nn5_760_submission_pyboost = pd.read_csv('/kaggle/input/copy-of-pyboost-secret-grandmaster-s-to-1d68b4/submission.csv', index_col='id')\nn6_776_submission_nn = pd.read_csv('/kaggle/input/4th-place-neural-net/submission.csv', index_col='id')\nn7_792_submission_lgbm = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')\nn8_813_submission_nlp = pd.read_csv('/kaggle/input/nlp-regression/submission.csv', index_col='id')\nn3_750_result_9 = pd.read_csv('/kaggle/input/2-pla-submit/result (9).csv', index_col='id')\nn4_767_result_8 = pd.read_csv('/kaggle/input/2-pla-submit/result (8).csv', index_col='id')\nn2_758_result_15 = pd.read_csv('/kaggle/input/2-pla-submit/result (15).csv', index_col='id')\nn9_736_3rd_place_psevdo_submission = pd.read_csv('/kaggle/input/3rd-place/3rd place psevdo submission(14).csv', index_col='id')\nn12_746_OP_pub530_myModels_exp1 = pd.read_csv('/kaggle/input/4-pl-sub-late/OP_pub530_myModels_exp1.csv', index_col='id')\nimport_732 = pd.read_csv('/kaggle/input/3rd-place/3rd place submission.csv', index_col='id')\nimport_746 = pd.read_csv('/kaggle/input/4-pl-sub-late/4th_Place_Ensembling_submission.csv', index_col='id')\nn13_729_1_pl = pd.read_csv('/kaggle/input/submit-best/submission.csv', index_col='id')\nn14_719_sub_b = pd.read_csv('/kaggle/input/1st-place-solution/submission.csv', index_col='id')\nimport_670 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729.csv', index_col='id')\nimport_657 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785.csv', index_col='id')\nimport_626 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add.csv', index_col='id')\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')\nimport_600 = pd.read_csv('/kaggle/input/op-0-670/submission732_some_746_some760_750_558_776_736_719_729_784_834_785_add_mult.csv', index_col='id')\nm1_792_4th_submission = pd.read_csv('/kaggle/input/lgbm-with-gene-aggregation-4th-place-writeup/submission.csv', index_col='id')\nm2_784_LS_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_NK.csv', index_col='id')\nm3_834_LS_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_RG.csv', index_col='id')\nm4_775_LS_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T4.csv', index_col='id')\nm5_959_LS_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/LS_T8.csv', index_col='id')\nm6_784_SVD_NK = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_NK.csv', index_col='id')\nm7_834_SVD_RG = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_RG.csv', index_col='id')\nm8_775_SVD_T4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T4.csv', index_col='id')\nm9_959_SVD_T8 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_T8.csv', index_col='id')\nm10_773_SVD_b_nk_m_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_nk_m_t4.csv', index_col='id')\nm11_786_SVD_b_t4_m_nk = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_b_t4_m_nk.csv', index_col='id')\nm12_768_SVD_nk_t4 = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4.csv', index_col='id')\nm13_771_SVD_nk_t4_t8_rg = pd.read_csv('/kaggle/input/24th-using-linear-algebra-priv-pub-0-768-0-582/SVD_nk_t4_t8_rg.csv', index_col='id')\nm14_746_op_late_submission = pd.read_csv('/kaggle/input/op-late-sub-pub530-mymodels-exp1/submission.csv', index_col='id')\nm15_863_plase30_submission = pd.read_csv('/kaggle/input/30-op2scp-transformer-deberta-v3-small-demo/submission.csv', index_col='id')\nm17_769_submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50 = pd.read_csv('/kaggle/input/pyboost-u900/submission_Pyboost_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')\nm18_790_submission_calibrated_max_d = pd.read_csv('/kaggle/input/pyboost-u900/submission_calibrated_max_depth10_ntrees5000_lr001_subsample1_colsample035_n_components50.csv', index_col='id')\nM19_785_13th_place_mlp = pd.read_csv('/kaggle/input/13th-place-solution-mlp/submission_example_MLP.csv', index_col='id')\nimport_755_10pl = pd.read_csv('/kaggle/input/10th-place/sub_submission.csv', index_col='id')\nimport_748_26pl = pd.read_csv('/kaggle/input/op-26-place/submission.csv', index_col='id')\nimport_577 = pd.read_csv('/kaggle/input/op-0-670/submi_mu_0.577.csv', index_col='id')\n\nThank you for taking the time to read. Feel free to ask any questions you may have.",
    "2558915": "Great work!! Congratulations 🎉"
  },
  "source": "meta"
}