Classification of Characteristic Factors of Brain Functional Network Using Decision Tree

Abstract

In order to explore the classification methods from imaging data, we propose to establish the feature classification order by building a classification tree with characteristic factors of brain functional network aided by calculating of network topology based on graph theory. By preprocessing the imaging data and modeling the networks with professional platform, we get the topology parameters as the input of the C4.5 decision tree, and then get the classification feature and classification tree by calculating the information gain ratio. The results show that the classification prefers holistic characteristics and quality characteristics. By programming and the samples test, it is feasible to extend the classification with network characters learning based on the processing results of multi-modal imaging data and machine learning algorithms.