A employment mobility characteristics study of college graduates based on neural networks

This study is based on national employment survey data of college graduates by Peking university in 2009. For the highly complex, nonlinear and uncertain characteristics in social science research, the study make use of neural network fitting simulation method, to better isolate the impact degree of gender differences in job characteristics and mobility characteristics in China. The results show that: Mobility characteristics of graduates have more significant gender differences: male graduates have a greater mobility than female. Except for job mobility, gender differences in schooling mobility, graduation mobility is 19.5% and 23.77%, gender differences in full mobility is reached 40.57%. Gender differences in job mobility is only 2.77 percentage point, this is mainly caused by female\´s "mobility inertia".