A Prediction for Classification of Highly Imbalanced Medical Dataset Using Databoost.IM with SVM

Recently, Class imbalance problems have growing interest because of their classification difficulty caused by the imbalanced class distributions. In particular, many ensemble learning and machine learning methods have been proposed for classification of imbalance problem. However, these methods producing poor predictive accuracy of classification for two-class imbalanced dataset. In this paper, we propose a new approach that combines an ensemblebased learning algorithm (DataBoost.IM) with Machine learning algorithm (SVM) to improve the predictive power of classifiers for imbalanced Liver data sets consisting of two classes. In the DataBoost.IM–SVM method identified accuracy of both the majority and minority classes from imbalanced liver datasets during execution. This method was evaluated by F-measures, G-mean and overall accuracy, against imbalanced data sets. Our results compares with other existing algorithm for imbalanced Liver data set.