Meihong Zhao and  Guangsheng Zhang. The prediction model of migrant workers’ returning home and Entrepreneurship Based on genetic algorithm and support vector machine. Dynamic Systems and Applications 29 (2020) No. 4, 1213 – 1226

https://doi.org/10.46719/dsa202029415

ABSTRACT.
75% of animals’ perception of the surrounding things is obtained through their own visual system. In the same way, vision system is also regarded as a tool for robot to sense the surrounding things, and it is also the premise for robot to move autonomously. This paper introduces the GA-SVM optimization algorithm based on the combination of genetic algorithm (GA) and support vector machine algorithm (SVM). The prediction ability of the algorithm is very strong. According to the 2016 China Statistical Yearbook statistics, nine factors are selected as the influencing factors, and the GA-SVM optimization algorithm is applied to the prediction of migrant workers’ returning home and entrepreneurship, and based on the idea and method of data modeling, Matlab is applied Simulation experiment. The results show that the average fitness value of GA-SVM is close to the optimal fitness value, indicating that each individual in the population is near the optimal solution, and the effect is good. The RMSE of GA-SVM is 364.062, and the root mean square error of GA-SVM is larger than this value, which shows that the application effect of GA-SVM is significantly better than that of SVM using cross validation method, and it is more suitable for precision prediction, and it also reflects that the optimization of SVM parameters by GA is reliable. This algorithm can be better applied in the prediction of migrant workers’ returning home and entrepreneurship, and can provide some help and support for the government in the policy reform of migrant workers’ entrepreneurship.

Keywords: support vector machine; genetic algorithm; migrant workers; entrepreneurship; prediction model