Now, without filtering out most of the features, our performance is 0.468237 on the left-out data, which is relatively not good compared to the leaderboard on Kaggle now, where the top 10 are all below 0.32
[0] train-rmsle:2.97083 test-rmsle:2.983
Multiple eval metrics have been passed: 'test-rmsle' will be used for early stopping.
Now, without filtering out most of the features, our performance is 0.468237 on the left-out data, which is relatively not good compared to the leaderboard on Kaggle now, where the top 10 are all below 0.32
[0] train-rmsle:2.97083 test-rmsle:2.983 Multiple eval metrics have been passed: 'test-rmsle' will be used for early stopping.
Will train until test-rmsle hasn't improved in 50 rounds. [5] train-rmsle:1.3304 test-rmsle:1.34331 [10] train-rmsle:0.882619 test-rmsle:0.901277 [15] train-rmsle:0.667358 test-rmsle:0.691634 [20] train-rmsle:0.552201 test-rmsle:0.581514 [25] train-rmsle:0.490082 test-rmsle:0.523132 [30] train-rmsle:0.456512 test-rmsle:0.493189 [35] train-rmsle:0.437507 test-rmsle:0.478101 [40] train-rmsle:0.427903 test-rmsle:0.470916 [45] train-rmsle:0.422715 test-rmsle:0.467869 [50] train-rmsle:0.41999 test-rmsle:0.466666 [55] train-rmsle:0.418106 test-rmsle:0.466477 [60] train-rmsle:0.416868 test-rmsle:0.466856 [65] train-rmsle:0.415938 test-rmsle:0.467272 [70] train-rmsle:0.414363 test-rmsle:0.467484 [75] train-rmsle:0.41295 test-rmsle:0.467877 [80] train-rmsle:0.411823 test-rmsle:0.468011 [85] train-rmsle:0.410196 test-rmsle:0.468204 [90] train-rmsle:0.408451 test-rmsle:0.468278 [95] train-rmsle:0.4069 test-rmsle:0.468237