摘要：Support Vector Machines (SVM) has shown a range of promising applications on classification problems. In this paper, we propose the genetic algorithm that employs Self-Adaptive Mutation Rate (SAMR) to develop kernel functions for SVM classifiers. The proposed SAMR model implemented the hybrid model for three advanced non-linear classification algorithms and shows competitive results in comparing to Grid SVM. Five publicly available datasets, cross validation correctness for Area Under Curve (AUC) have been involved. Improvements achieved may lead to biomarkers results
