AN IMPROVED RANDOM FOREST ALGORITHM FOR A CLASS IMBALANCE CLASSIFICATION FOR MEDICAL DIAGNOSTIC

This paper proposes an improved random forest algorithm for classifying diabetic data. This algorithm is particularly designed for analyzing very high dimensional data with multiple classes whose well-known representative data. A novel feature early classification of classes for subspace sampling and tree selection method are developed and synergistically served for making random forest framework well suited to categorize diabetic data. With the new feature early classification of classes for subspace sampling and tree selection method, we can effectively reduce subspace size and improve classification performance without increasing error bound and parallel implementation of improved random forest algorithm. We apply the proposed method on diabetic data sets with 8 characteristics. The results have demonstrated that this improved random forests outperformed the original Random Forest methods in terms of classification performance.