Class imbalance problem has been a challenging research problem in the fieldsof machine learning and data mining as most real life datasets are imbalanced.Several existing machine learning algorithms try to maximize the accuracyclassification by correctly identifying majority class samples while ignoringthe minority class. However, the concept of the minority class instancesusually represents a higher interest than the majority class. Recently, severalcost sensitive methods, ensemble models and sampling techniques have been usedin literature in order to classify imbalance datasets. In this paper, wepropose MEBoost, a new boosting algorithm for imbalanced datasets. MEBoostmixes two different weak learners with boosting to improve the performance onimbalanced datasets. MEBoost is an alternative to the existing techniques suchas SMOTEBoost, RUSBoost, Adaboost, etc. The performance of MEBoost has beenevaluated on 12 benchmark imbalanced datasets with state of the art ensemblemethods like SMOTEBoost, RUSBoost, Easy Ensemble, EUSBoost, DataBoost.Experimental results show significant improvement over the other methods and itcan be concluded that MEBoost is an effective and promising algorithm to dealwith imbalance datasets. The python version of the code is available here:https://github.com/farshidrayhanuiu/
展开▼
