Research on Recognition of Medical Insurance Fraud Based on Modified Support Vector Machine
Yifeng Dou, Hailing Xiong · 2017 International Conference on Computer Technology, Electronics and Communication (ICCTEC) · 2017
With the rapid development of China's medical insurance industry, fraud irregularities are also constantly increasing with renovation of forms and means. Then the way of detecting the fraud automatically and efficiently has become the major issue on which much attention is paid. This paper selects genetic algorithm (GA) and particle swarm optimization (PSO) as the basic algorithm. Spontaneously, in order to comprehensively utilize the local search space of GA and swift convergence capability of PSO algorithm as well as improve the poor local search of PSO by introducing simulated annealing (SA), this paper proposes an algorithm that utilizes three classical algorithms to optimize support vector machine (GASAPSO-SVM) and ultimately verifies its algorithm on the medical insurance fraud dataset.