Tuning a Multiple Classifier System for Side Effect Discovery Using Genetic Algorithms

摘要: In previous work, a novel supervised framework implementing binary classifier was presented that obtained excellent results for side effect discovery. Interestingly, unique effects were identified when different classifiers used within the framework, prompting investigation of applying multiple system. this paper we investigate tuning classifying system using genetic algorithms. The research show trained algorithms can obtain higher partial area under receiver operating characteristic curve than single classifier. Furthermore, is able to detect efficiently and obtains low false positive rate.