Core Classifier Algorithm: A Hybrid Classification Algorithm Based on Class Core and Clustering

Machine learning classification algorithms vary drastically in their approaches, and researchers have always been trying to reduce the common boundaries of nonlinear classification, overlapping, or noise. This study summarizes steps hybridizing a new algorithm named Core Classify Algorithm (CCA) derived from K-nearest neighbor (KNN) an unsupervised partitioning (K-means), aiming avoid unrepresentative Cores clusters while finding similarities. hybridization step is meant harvest benefits combining two by changing results through iteration obtain most optimal classifying data according labels with more higher accuracy better computational efficiency. Our approach was tested on total five datasets different domains: one phishing URL, three healthcare, synthetic dataset. demonstrate that CCA model non-linear experiments representing lower than dataset which represented linear achieved 100%, equal rank Random Forest, Support Vector Machine, Decision Trees. Moreover, our also can be used exploit flaws specific further improve performance.