Using Simple Gaussian Mixture Model for Multi-Class Classification Based on Tumor Gene Expression Data

In this paper, we developed a novel multi-class classification method combining the ideal of discriminant analysis and Gaussian Mixture Model. Different from binary classification, this method reserves more information and is useful for multi-class tumor subtypes diagnosis and treatment. Four datasets, ALL-AML-3, ALL-AML-3, MLL and ALL, were collected and used to evaluate the prediction performance. The classification accuracies are all about 2.5% higher than KNN classifier and comparable well to SVM for leave-one-out cross validation. The results demonstrate that this method is simple and efficient even more less computational cost. It is a useful tool for multi-class tumor classification.