{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"

Tabular Playground Series - Nov 2021⚡

\n
","metadata":{}},{"cell_type":"markdown","source":"\n\n## 🔷 Introduction\n\nThis dataset is synthetic, but based on a real dataset and generated using a CTGAN. The original dataset deals with predicting identifying spam emails via various extracted features from the email. Although the features are anonymized, they have properties relating to real-world features. The dataset has 100 features and the response variable is a binary variable.\n\nIn my first notebook I have fitted linear models, tree based models and boosting models. The highest accuracy was given by linear models. So in this notebook I will try with linear models to improve accuracy more.","metadata":{}},{"cell_type":"markdown","source":"\n\n#### 🎭 loading.. dataset","metadata":{}},{"cell_type":"code","source":"#importing librariaes\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:09.734839Z","iopub.execute_input":"2021-11-30T13:51:09.735198Z","iopub.status.idle":"2021-11-30T13:51:10.649798Z","shell.execute_reply.started":"2021-11-30T13:51:09.735097Z","shell.execute_reply":"2021-11-30T13:51:10.649188Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n for filename in filenames:\n print(os.path.join(dirname, filename))\nimport warnings\nwarnings.filterwarnings(\"ignore\")","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:10.651217Z","iopub.execute_input":"2021-11-30T13:51:10.651555Z","iopub.status.idle":"2021-11-30T13:51:10.660193Z","shell.execute_reply.started":"2021-11-30T13:51:10.651526Z","shell.execute_reply":"2021-11-30T13:51:10.659107Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train = pd.read_csv('/kaggle/input/tabular-playground-series-nov-2021/train.csv')\ntest = pd.read_csv('/kaggle/input/tabular-playground-series-nov-2021/test.csv')\n\nprint(train.shape)\nprint(test.shape)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:10.663409Z","iopub.execute_input":"2021-11-30T13:51:10.66363Z","iopub.status.idle":"2021-11-30T13:51:39.152784Z","shell.execute_reply.started":"2021-11-30T13:51:10.663602Z","shell.execute_reply":"2021-11-30T13:51:39.151903Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train.head()","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:39.154301Z","iopub.execute_input":"2021-11-30T13:51:39.154818Z","iopub.status.idle":"2021-11-30T13:51:39.189622Z","shell.execute_reply.started":"2021-11-30T13:51:39.15477Z","shell.execute_reply":"2021-11-30T13:51:39.189135Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train.dtypes","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:39.191132Z","iopub.execute_input":"2021-11-30T13:51:39.19147Z","iopub.status.idle":"2021-11-30T13:51:39.198352Z","shell.execute_reply.started":"2021-11-30T13:51:39.191436Z","shell.execute_reply":"2021-11-30T13:51:39.197494Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train=train.drop(['id'],1)\ntest=test.drop(['id'],1)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:39.199402Z","iopub.execute_input":"2021-11-30T13:51:39.199671Z","iopub.status.idle":"2021-11-30T13:51:39.538265Z","shell.execute_reply.started":"2021-11-30T13:51:39.199645Z","shell.execute_reply":"2021-11-30T13:51:39.537562Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from collections import Counter\nprint(sorted(Counter(train['target']).items()))\nsns.countplot(train['target'],palette='OrRd')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:39.539349Z","iopub.execute_input":"2021-11-30T13:51:39.539561Z","iopub.status.idle":"2021-11-30T13:51:39.885374Z","shell.execute_reply.started":"2021-11-30T13:51:39.539536Z","shell.execute_reply":"2021-11-30T13:51:39.8846Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sns.heatmap(train.corr())","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:39.886958Z","iopub.execute_input":"2021-11-30T13:51:39.887246Z","iopub.status.idle":"2021-11-30T13:51:56.907242Z","shell.execute_reply.started":"2021-11-30T13:51:39.887208Z","shell.execute_reply":"2021-11-30T13:51:56.906443Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sns.heatmap(train.isnull(),yticklabels=False,cbar=False)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:51:56.908387Z","iopub.execute_input":"2021-11-30T13:51:56.908591Z","iopub.status.idle":"2021-11-30T13:53:06.845653Z","shell.execute_reply.started":"2021-11-30T13:51:56.908564Z","shell.execute_reply":"2021-11-30T13:53:06.844818Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_x=train.drop(['target'],1)\ntest_x=test","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:06.846742Z","iopub.execute_input":"2021-11-30T13:53:06.847018Z","iopub.status.idle":"2021-11-30T13:53:06.991679Z","shell.execute_reply.started":"2021-11-30T13:53:06.846983Z","shell.execute_reply":"2021-11-30T13:53:06.990814Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X= train_x\ny= train['target']","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:06.992962Z","iopub.execute_input":"2021-11-30T13:53:06.993259Z","iopub.status.idle":"2021-11-30T13:53:07.009333Z","shell.execute_reply.started":"2021-11-30T13:53:06.99322Z","shell.execute_reply":"2021-11-30T13:53:07.008632Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#dividing the dataset into train and test sets\nfrom sklearn.model_selection import train_test_split\nx_train,x_test,y_train,y_test=train_test_split(X,y,test_size=0.20,random_state=1234)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:07.011976Z","iopub.execute_input":"2021-11-30T13:53:07.012449Z","iopub.status.idle":"2021-11-30T13:53:07.743137Z","shell.execute_reply.started":"2021-11-30T13:53:07.01239Z","shell.execute_reply":"2021-11-30T13:53:07.742377Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n## 🔷 sampling to reduce training time\n\n","metadata":{}},{"cell_type":"code","source":"x_train2,x_test2,y_train2,y_test2=train_test_split(x_train,y_train,test_size=0.15,random_state=1234)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:07.744263Z","iopub.execute_input":"2021-11-30T13:53:07.74453Z","iopub.status.idle":"2021-11-30T13:53:08.195682Z","shell.execute_reply.started":"2021-11-30T13:53:07.744497Z","shell.execute_reply":"2021-11-30T13:53:08.195028Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"x_train3,x_test3,y_train3,y_test3=train_test_split(x_train2,y_train2,test_size=0.14,random_state=1234)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:08.198555Z","iopub.execute_input":"2021-11-30T13:53:08.19881Z","iopub.status.idle":"2021-11-30T13:53:08.555966Z","shell.execute_reply.started":"2021-11-30T13:53:08.198779Z","shell.execute_reply":"2021-11-30T13:53:08.555269Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"x_train4,x_test4,y_train4,y_test4=train_test_split(x_train3,y_train3,test_size=0.15,random_state=1234)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:08.559115Z","iopub.execute_input":"2021-11-30T13:53:08.559305Z","iopub.status.idle":"2021-11-30T13:53:08.881517Z","shell.execute_reply.started":"2021-11-30T13:53:08.559281Z","shell.execute_reply":"2021-11-30T13:53:08.880546Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"x_train5,x_test5,y_train5,y_test5=train_test_split(x_train4,y_train4,test_size=0.04,random_state=1234)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:08.882675Z","iopub.execute_input":"2021-11-30T13:53:08.882901Z","iopub.status.idle":"2021-11-30T13:53:09.132601Z","shell.execute_reply.started":"2021-11-30T13:53:08.882873Z","shell.execute_reply":"2021-11-30T13:53:09.131747Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"x_train1,x_test1,y_train1,y_test1=train_test_split(x_train5,y_train5,test_size=0.1,random_state=1234)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:09.133939Z","iopub.execute_input":"2021-11-30T13:53:09.134157Z","iopub.status.idle":"2021-11-30T13:53:09.380557Z","shell.execute_reply.started":"2021-11-30T13:53:09.13413Z","shell.execute_reply":"2021-11-30T13:53:09.379841Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(x_train1.shape)\nprint(x_test.shape)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:09.381739Z","iopub.execute_input":"2021-11-30T13:53:09.382072Z","iopub.status.idle":"2021-11-30T13:53:09.388011Z","shell.execute_reply.started":"2021-11-30T13:53:09.382032Z","shell.execute_reply":"2021-11-30T13:53:09.387078Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n## ⌛ Model fitting","metadata":{}},{"cell_type":"code","source":"from sklearn.linear_model import LogisticRegression\nfrom sklearn.discriminant_analysis import LinearDiscriminantAnalysis\nfrom sklearn.tree import DecisionTreeClassifier \nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.ensemble import GradientBoostingClassifier\nfrom xgboost import XGBClassifier\nfrom sklearn.ensemble import ExtraTreesClassifier\nfrom sklearn.ensemble import BaggingClassifier\nfrom sklearn.ensemble import AdaBoostClassifier\nfrom sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n\n#libraries for model evaluation\nfrom sklearn.metrics import accuracy_score\nfrom sklearn.metrics import plot_confusion_matrix\nfrom sklearn.metrics import classification_report","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:09.389193Z","iopub.execute_input":"2021-11-30T13:53:09.389699Z","iopub.status.idle":"2021-11-30T13:53:09.656764Z","shell.execute_reply.started":"2021-11-30T13:53:09.389669Z","shell.execute_reply":"2021-11-30T13:53:09.655942Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from sklearn.linear_model import RidgeClassifier\nrc =RidgeClassifier()\nmodel0=rc.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model0.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model0.score(x_test,y_test))\nrcpred = rc.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for logistic regression\")\nprint(classification_report(rcpred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for logistic regression\")\ndisplr = plot_confusion_matrix(rc, x_test, y_test,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:09.658111Z","iopub.execute_input":"2021-11-30T13:53:09.658341Z","iopub.status.idle":"2021-11-30T13:53:11.403094Z","shell.execute_reply.started":"2021-11-30T13:53:09.658313Z","shell.execute_reply":"2021-11-30T13:53:11.402174Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#logistic regression\nlr = LogisticRegression(max_iter=10000,penalty='l2')\nmodel1=lr.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model1.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model1.score(x_test,y_test))\nlrpred = lr.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for logistic regression\")\nprint(classification_report(lrpred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for logistic regression\")\ndisplr = plot_confusion_matrix(lr, x_test, y_test,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:53:11.404582Z","iopub.execute_input":"2021-11-30T13:53:11.404921Z","iopub.status.idle":"2021-11-30T13:55:48.303712Z","shell.execute_reply.started":"2021-11-30T13:53:11.404878Z","shell.execute_reply":"2021-11-30T13:55:48.30317Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#linear discriminant analysis\nlda = LinearDiscriminantAnalysis()\nmodel2=lda.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model2.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model2.score(x_test,y_test))\n\nldapred = lda.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for linear discriminant analysis\")\nprint(classification_report(ldapred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for linear discriminant analysis\")\ndisplr = plot_confusion_matrix(lda, x_test, y_test ,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:55:48.305202Z","iopub.execute_input":"2021-11-30T13:55:48.306014Z","iopub.status.idle":"2021-11-30T13:55:54.344578Z","shell.execute_reply.started":"2021-11-30T13:55:48.305963Z","shell.execute_reply":"2021-11-30T13:55:54.343915Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#quadratic discriminant analysis\nqda = QuadraticDiscriminantAnalysis()\nmodel3=qda.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model3.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model3.score(x_test,y_test))\n\nqdapred = qda.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for linear discriminant analysis\")\nprint(classification_report(qdapred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for linear discriminant analysis\")\ndisplr = plot_confusion_matrix(qda, x_test, y_test ,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:55:54.345627Z","iopub.execute_input":"2021-11-30T13:55:54.346221Z","iopub.status.idle":"2021-11-30T13:56:00.034892Z","shell.execute_reply.started":"2021-11-30T13:55:54.346183Z","shell.execute_reply":"2021-11-30T13:56:00.034079Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
","metadata":{}},{"cell_type":"code","source":"# stacking classifier \nfrom sklearn.ensemble import StackingClassifier\nfrom sklearn.pipeline import make_pipeline\nfrom sklearn.preprocessing import StandardScaler\n\nestimators = [('rf',make_pipeline(StandardScaler(),LinearDiscriminantAnalysis()) ),('ext', make_pipeline(StandardScaler(),LogisticRegression(max_iter=10000,penalty='l2')))]\nsc= StackingClassifier( estimators=estimators)\n\nmodel5=sc.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model5.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model5.score(x_test,y_test))\n\nscpred = sc.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for voting classifier\")\nprint(classification_report(scpred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for voting classifier\")\ndisplr = plot_confusion_matrix(sc, x_test, y_test ,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:56:00.036965Z","iopub.execute_input":"2021-11-30T13:56:00.037644Z","iopub.status.idle":"2021-11-30T13:56:39.714354Z","shell.execute_reply.started":"2021-11-30T13:56:00.037594Z","shell.execute_reply":"2021-11-30T13:56:39.713559Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from catboost import CatBoostClassifier\n\ncc = CatBoostClassifier(silent=True )\nmodel6=cc.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model6.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model6.score(x_test,y_test))\n\nccpred = cc.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for extra tree classifier\")\nprint(classification_report(ccpred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for extra tree classifier\")\ndisplr = plot_confusion_matrix(cc, x_test, y_test ,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:56:39.71566Z","iopub.execute_input":"2021-11-30T13:56:39.716016Z","iopub.status.idle":"2021-11-30T13:57:50.821483Z","shell.execute_reply.started":"2021-11-30T13:56:39.715965Z","shell.execute_reply":"2021-11-30T13:57:50.820603Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n## ⏰ tuning of logistic regression","metadata":{}},{"cell_type":"code","source":"from sklearn.model_selection import GridSearchCV\nLR = LogisticRegression()\nLRparam_grid = {\n 'C': [ 100,250, 400, 500,1000],\n 'penalty': ['l1', 'l2'],\n 'max_iter': [200,400,600],\n 'solver': ['liblinear']\n}\nLR_search = GridSearchCV(LR, param_grid=LRparam_grid, refit = True, verbose = 3, cv=3)\n\n# fitting the model for grid search \nLR_search.fit(x_train1 , y_train1)\nLR_search.best_params_\n# summarize\nprint('Mean Accuracy: %.3f' % LR_search.best_score_)\nprint('Config: %s' % LR_search.best_params_)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T13:57:50.823115Z","iopub.execute_input":"2021-11-30T13:57:50.82351Z","iopub.status.idle":"2021-11-30T14:23:26.276293Z","shell.execute_reply.started":"2021-11-30T13:57:50.823465Z","shell.execute_reply":"2021-11-30T14:23:26.274759Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#logistic regression\nlr = LogisticRegression(max_iter=600,penalty='l1',C=400,solver='liblinear')\nmodel7=lr.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model7.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model7.score(x_test,y_test))\nlrpred = lr.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for logistic regression\")\nprint(classification_report(lrpred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for logistic regression\")\ndisplr = plot_confusion_matrix(lr, x_test, y_test,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T14:23:26.278119Z","iopub.status.idle":"2021-11-30T14:23:26.278637Z","shell.execute_reply.started":"2021-11-30T14:23:26.278389Z","shell.execute_reply":"2021-11-30T14:23:26.278416Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Let's see stacking classifier with tuned logistic and linear discriminant analysis models.","metadata":{}},{"cell_type":"code","source":"# stacking classifier \nfrom sklearn.ensemble import StackingClassifier\nfrom sklearn.pipeline import make_pipeline\nfrom sklearn.preprocessing import StandardScaler\n\nestimators = [('lda',make_pipeline(StandardScaler(),LinearDiscriminantAnalysis() )),('lr', make_pipeline(StandardScaler(),LogisticRegression(max_iter=600,penalty='l1',C=400, solver='liblinear')))]\nscfinal= StackingClassifier( estimators=estimators)\n\nmodel8=scfinal.fit(x_train1, y_train1)\nprint(\"train accuracy:\",model8.score(x_train1, y_train1),\"\\n\",\"test accuracy:\",model8.score(x_test,y_test))\n\nscpred = scfinal.predict(x_test)\nprint(\"\\n\")\nprint(\"classification report for voting classifier\")\nprint(classification_report(scpred,y_test))\nprint(\"\\n\")\nprint(\"confusion matrix for voting classifier\")\ndisplr = plot_confusion_matrix(scfinal, x_test, y_test ,cmap=plt.cm.Blues , values_format='d')","metadata":{"execution":{"iopub.status.busy":"2021-11-30T14:23:26.280231Z","iopub.status.idle":"2021-11-30T14:23:26.280779Z","shell.execute_reply.started":"2021-11-30T14:23:26.280516Z","shell.execute_reply":"2021-11-30T14:23:26.280542Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"predictions=model8.predict_proba(test)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T14:23:26.282378Z","iopub.status.idle":"2021-11-30T14:23:26.282832Z","shell.execute_reply.started":"2021-11-30T14:23:26.282583Z","shell.execute_reply":"2021-11-30T14:23:26.282606Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_submission = pd.read_csv('/kaggle/input/tabular-playground-series-nov-2021/sample_submission.csv')\ndf_submission['target']=predictions[:,1]\ndf_submission.to_csv('submission.csv',index=False)","metadata":{"execution":{"iopub.status.busy":"2021-11-30T14:23:26.284184Z","iopub.status.idle":"2021-11-30T14:23:26.284757Z","shell.execute_reply.started":"2021-11-30T14:23:26.284495Z","shell.execute_reply":"2021-11-30T14:23:26.28452Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n## 💎 Results","metadata":{}},{"cell_type":"markdown","source":" In this notebook I used a subset of the training sample randomly to fit the models. It gives better results than my previous notebook. In that also linear models and catboost classifiers had the best performance and tree based models tend to overfitting. Here logistic regression model and the linear discriminant analysis was tuned using a grid search. The highest accuracy achieved is 74.01%which is a increase of 3% than previous effort.","metadata":{}},{"cell_type":"markdown","source":"\n
","metadata":{}},{"cell_type":"markdown","source":" Any suggestions,questions or feedback regarding this notebook are highly appreciated. THANK YOU 😊 ","metadata":{}}]}