ENHANCING THE EFFECTIVENESS AND ACCURACY OF GENERALIZED INSTANCES OVER IMBALANCED PROBLEM USING ML

In Machine Learning (ML), Classification with imbalanced datasets is considered to be a newchallenge for researches in the framework of data mining. The imbalance problem occurs in manyexamples that represents one of the classes of the dataset is much lower than the other classes. Totackle with imbalance problem, pre-processing the datasets applied with oversampling method(SMOTE) was previously proposed. Generalized instances are belonging to the family of NestedGeneralized Exemplar, which achieves storing objects in Euclidean n-space. The mostrepresentative mode used in NGE learning is: classical-BNGE and RISE, recent-INNER, ruleinduction-RIPPER and PART. The Fuzzy Neural Network approach, which is a combination offuzzy logic and neural networks and called as Neuro Fuzzy System, which could improve theperformance and accuracy of the existing system. The proposed approach deals with thecomparison of NGE learning without using SMOTE methods.