Hackerearth machine learning challenge - 6 (https://www.hackerearth.com/challenge/competitive/machine-learning-challenge-6-1/), 3rd rank solution (https://www.hackerearth.com/challenge/competitive/machine-learning-challenge-6-1/leaderboard/)
Performance of models:
| Model | Score | calibrated using sklearn |
|---|---|---|
| catboost | 0.7488 | |
| lgbm | 0.76723 | 0.7674 |
| etree | 0.6202 | |
| knn | 0.5514 | |
| lgb_without_munc | 0.76663 | 0.76686 |
| lgb_without_ward | 0.7662 | 0.76719 |
| catb_without_munc | 0.76201 | |
| catb_averaged | 0.76558 | |
| nnet | 0.7355 | |
| catb_less_feat | 0.76821 | 0.76302 |
Best_score=prob_lgb(0.3)+prob_lgb_munc(0.2)+prob_lgb_ward(0.1)+prob_catb_less_feat(0.4)
0.76978