{
  "id": 518955,
  "title": "166th simple solution",
  "url": "/competitions/leash-BELKA/discussion/518955",
  "author_name": "CodeHacker",
  "post_date": "2024-07-09T03:32:52.809000",
  "votes": 2,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I use two data format. </p>\n<ol>\n<li>BELKA: Shrinking the dataset (<a href=\"https://www.kaggle.com/code/shlomoron/belka-shrinking-the-dataset\" target=\"_blank\">https://www.kaggle.com/code/shlomoron/belka-shrinking-the-dataset</a>)</li>\n<li>Just change molecules to numbers (<a href=\"https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data\" target=\"_blank\">https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data</a>)</li>\n</ol>\n<p>Using 1d-cnn 1 get two solution csv files. And using BELKA: Shrinking the dataset I got 3 three solution.</p>\n<p>Using 1d-cnn, i do 5 fold which same as <a href=\"https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data\" target=\"_blank\">https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data</a>. The other one is little change of previous noteboot.</p>\n<p>I more focus on BELKA: Shrinking the dataset case which i don't  know that it is worse data for me.</p>\n<p>I using buildingblock1_smiles, buildingblock2_smiles, buildingblock3_smiles are input.<br>\nAnd first notebook is only using dense network. It gives me LB 0.366.<br>\nI add the residual one from previous. It gives me LB 0.446.</p>\n<p>My result can see below image.<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13540117%2F9bdb10de244449f237aaf5db7b0f90b6%2F2024-07-09%20122227.png?generation=1720495871273553&amp;alt=media\"></p>",
  "messages": [
    {
      "id": 2912704,
      "postDate": "2024-07-09T03:32:52.810Z",
      "content": "<p>I use two data format. </p>\n<ol>\n<li>BELKA: Shrinking the dataset (<a href=\"https://www.kaggle.com/code/shlomoron/belka-shrinking-the-dataset\" target=\"_blank\">https://www.kaggle.com/code/shlomoron/belka-shrinking-the-dataset</a>)</li>\n<li>Just change molecules to numbers (<a href=\"https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data\" target=\"_blank\">https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data</a>)</li>\n</ol>\n<p>Using 1d-cnn 1 get two solution csv files. And using BELKA: Shrinking the dataset I got 3 three solution.</p>\n<p>Using 1d-cnn, i do 5 fold which same as <a href=\"https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data\" target=\"_blank\">https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data</a>. The other one is little change of previous noteboot.</p>\n<p>I more focus on BELKA: Shrinking the dataset case which i don't  know that it is worse data for me.</p>\n<p>I using buildingblock1_smiles, buildingblock2_smiles, buildingblock3_smiles are input.<br>\nAnd first notebook is only using dense network. It gives me LB 0.366.<br>\nI add the residual one from previous. It gives me LB 0.446.</p>\n<p>My result can see below image.<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13540117%2F9bdb10de244449f237aaf5db7b0f90b6%2F2024-07-09%20122227.png?generation=1720495871273553&amp;alt=media\"></p>",
      "rawMarkdown": "I use two data format. \n\n1. BELKA: Shrinking the dataset (https://www.kaggle.com/code/shlomoron/belka-shrinking-the-dataset)\n2. Just change molecules to numbers (https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data)\n\nUsing 1d-cnn 1 get two solution csv files. And using BELKA: Shrinking the dataset I got 3 three solution.\n\nUsing 1d-cnn, i do 5 fold which same as https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data. The other one is little change of previous noteboot.\n\nI more focus on BELKA: Shrinking the dataset case which i don't  know that it is worse data for me.\n\nI using buildingblock1_smiles, buildingblock2_smiles, buildingblock3_smiles are input.\nAnd first notebook is only using dense network. It gives me LB 0.366.\nI add the residual one from previous. It gives me LB 0.446.\n\nMy result can see below image.\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13540117%2F9bdb10de244449f237aaf5db7b0f90b6%2F2024-07-09%20122227.png?generation=1720495871273553&alt=media)",
      "votes": 2
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2912704": "I use two data format. \n\n1. BELKA: Shrinking the dataset (https://www.kaggle.com/code/shlomoron/belka-shrinking-the-dataset)\n2. Just change molecules to numbers (https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data)\n\nUsing 1d-cnn 1 get two solution csv files. And using BELKA: Shrinking the dataset I got 3 three solution.\n\nUsing 1d-cnn, i do 5 fold which same as https://www.kaggle.com/code/ahmedelfazouan/belka-1dcnn-starter-with-all-data. The other one is little change of previous noteboot.\n\nI more focus on BELKA: Shrinking the dataset case which i don't  know that it is worse data for me.\n\nI using buildingblock1_smiles, buildingblock2_smiles, buildingblock3_smiles are input.\nAnd first notebook is only using dense network. It gives me LB 0.366.\nI add the residual one from previous. It gives me LB 0.446.\n\nMy result can see below image.\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13540117%2F9bdb10de244449f237aaf5db7b0f90b6%2F2024-07-09%20122227.png?generation=1720495871273553&alt=media)"
  }
}