phunterlau/kaggle_higgs

My winning solution for Kaggle Higgs Machine Learning Challenge (single classifier, xgboost)

kaggle_higgs

My winning solution for Kaggle Higgs Machine Learning Challenge (single classifier, xgboost)

The full description is linked to this page: http://no2147483647.wordpress.com/2014/09/17/winning-solution-of-kaggle-higgs-competition-what-a-single-model-can-do/

Thanks to xgboost for providing the fast GBM model. Thanks to glh3 for their original work on feature interface https://github.com/ghl3/higgs-kaggle

#usage

test_xgboost_pandas.py: the cross validation script. the original version is from Kaggle forum https://www.kaggle.com/c/higgs-boson/forums/t/8207/to-ams-3-6-model-can-you-share-you-local-cv-score/44825#post44825

higgs-numpy_pandas.py: the model generating script

higgs-pred_pandas.py: the submission generating script

run.all.sh: a simple script of generating model and generating submission

By running the default parameter of 3000 steps and max_depth of 9, one can reproduce the current best leaderboard score of 3.73 which is ranked 25th/1792

kaggle_higgs's Issues

`eta_plus(x, y)` error

def eta_plus(x, y):
    return np.sqrt(x*x/2 + 2*y*y - 2*x*y)

It looks as if there's a small error here, the right way is:

def eta_plus(x, y):
    return np.sqrt((x**2) / 2 + (y**2) / 2 - x*y)

A counterexample for your function is (x, y) = (0, 2) with a distance sqrt(2) to x=y