Cat Boost: unbiased boosting with categorical features

In this video we give a brief overview of the paper CatBoost: unbiased boosting with categorical features which has been accepted to the Thirtysecond Conference on Neural Information Processing Systems (NeurIPS, NIPS 2018). Here we describe one of the two paper s contributions, the ordered boosting, which was proposed to prevent prediction shift caused by a special kind of target leakage present in all currently existing implementations of gradient boosting algorithms. Find out more about CatBoost at h