Imbalanced Data Learning Approaches Review

Abstract

The present work deals with a well-known problem in machine learning, that classes have generally skewed prior probabilities distribution. This situation of imbalanced data is a handicap when trying to identify the minority classes , usually more interesting one In real world applications. This paper is an attempt to list the different approachs proposed in scientific research to deal with the imbalanced data learning, as well a comparison between various applications cases performed on this subject. KEYWORDS