Closed der-nico closed 7 years ago
In the implementation of AdaBoost the threshold value (to displace zero values) was set to the minimum float value, but it should be the float machine epsilon. See implementation in sklearn "_samme_proba" function: https://github.com/scikit-learn/scikit-learn/blob/ab93d65/sklearn/ensemble/weight_boosting.py#L288
Thanks for your contribution @der-nico!
Regards, Darius
In the implementation of AdaBoost the threshold value (to displace zero values) was set to the minimum float value, but it should be the float machine epsilon. See implementation in sklearn "_samme_proba" function: https://github.com/scikit-learn/scikit-learn/blob/ab93d65/sklearn/ensemble/weight_boosting.py#L288