Prediction of employment figures in the three main industries based on extreme learning machine

Xiaoyan Li, Min Zhang

Xi'an Technological University, Shaanxi, 710032, China

Extreme learning machine, with a fast speed of training, achieves globally optimal solutions and excellent generalization ability. This work is based on the production value and employment figure of the three main industries in China during 1996-2012 as research objects. An ELM employment figure prediction model was constructed with production value and the employment figure of the three main industries respectively as the input and output of extreme learning machine. Simulation test results proved a good effect and high accuracy of ELM employment figure prediction model. Besides, the comparison of ELM, BP and RBF algorithms further proved the effectiveness and precision of ELM, which has certain practical application value.