Higher School of Economics
| Project | Dataset | Features | Size | Scaled | Fillna | Zero | Algorithm | Cross-validation | Optimization | Metrics |
|---|---|---|---|---|---|---|---|---|---|---|
| titanic | Titanic | 4 | 891 (714) | no | remove | 9.38% | DecisionTreeClassifier | feature_importances | ||
| wine | Wine | 13 | 178 | scale | no | 0% | KNeighborsClassifier | KFold(5) | n_neighbors=1/51 | Accuracy |
| housing | Boston | 13 | 506 | no | no | 0% | KNeighborsRegressor | KFold(5) | p | MSE |
| perceptron | Perceptron | 2 | 300, 200 (test) | fit_transform/transform (StandardScaler) | no | 0% | Perceptron | train/test sets | Accuracy | |
| svm | SVM | 2 | 10 | no | no | 0% | SVC | |||
| svm-texts | 20 newsgroups | text (28382) | 1786 | fit_transform (TF/IDF) | no | 99.99% | SVC (TF/IDF) | KFold(5) | coef | |
| logistic | Logistic | 2 | 205 | no | no | 0% | LogisticRegression (own impl.) | AUC-ROC | ||
| metrics | 2 datasets | 200 | no | no | Accuracy, Precision, Recall, F1 | |||||
| linreg | CV/Salary | text (24627) | 60000 | fit_transform/transform (TF/IDF) | no | 99.99% | Ridge (TF/IDF) | train/test sets | ||
| pca | Close prices | 30 | 374 | no | no | 0% | PCA | close/djia sets | n_components=10 | corrcoef |
| forest | Abalone | 8 | 4177 | no | no | 4.02% | RandomForestRegressor | KFold(5) | n_estimators=1/51 | R2 |
| gbm | Bioresponse | 1776 | 3751 | no | no | 83.96% | GradientBoostingClassifier, RandomForestClassifier | train/test=0.8 | n_estimators | log-loss |
| clustering | Parrots (713x474) | 3 (RGB) | 337962 | no | no | 8.35% | KMeans | n_clusters=1/21 | PSNR, MSE | |
| dota2 | Dota 2 | 101 (+108 words bag) | 97230, 17177 (test) | scale | yes | 19.37% (49.49%) | GradientBoostingClassifier, LogisticRegression | KFold(5) | n_estimators, C | AUC-ROC |