Jaehyun Lee님의 게시글

질문&답변

피처별 회귀계수 시각화

오.. 교수님 코드를 그대로 실행해봤는데 모델 결과가 약간 다르게 나오긴하네요 ### RandomForestRegressor ### RMSLE: 0.355, RMSE: 50.466, MAE: 31.198 ### GradientBoostingRegressor ### RMSLE: 0.330, RMSE: 53.342, MAE: 32.750 ### XGBRegressor ### RMSLE: 0.339, RMSE: 51.475, MAE: 31.357 [LightGBM] [Info] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000588 seconds. You can set force_row_wise=true to remove the overhead. And if memory is not enough, you can set force_col_wise=true . [LightGBM] [Info] Total Bins 348 [LightGBM] [Info] Number of data points in the train set: 7620, number of used features: 72 [LightGBM] [Info] Start training from score 4.582043 ### LGBMRegressor ### RMSLE: 0.319, RMSE: 47.215, MAE: 29.029