FUSION OF HETEROGENEOUS FEATURES FOR MAJOR DEPRESSION DISORDER CLASSIFICATION BASED ON QDM-RANKED GENETIC ALGORITHM

摘要

Major depression disorder is a mental disorder which impacts various aspects of society. Fusion of heterogeneous features from different signal sources is a challenge task, especially with some flicking features different from persons or physiological conditions. The aim of this research is to develop a fusion model based on correlation, quartile discriminant measurement, and genetic algorithm. The result indicates heterogeneous features successfully fused for MDD classification. Finally, it provides 100% and 70% accuracy for training and testing datasets, respectively.