【摘要】 采用混合优化策略训练神经网络,进而实现地区人力资源数据的时间序列预测.神经网络,尤其是应用反向传播(back propagation,简称BP)算法训练的神经网络,被广泛应用于预测中.但是BP神经网络训练速度慢、容易陷入局部极值.遗传算法(genetic algorithm,简称GA)具有很好的全局寻优性.因而提出将BP和GA结合起来的混合优化策略训练神经网络,来实现人力资源数据预测.与BP算法相比,数值计算结果表明预测精度高、速度快,为地区人力资源数据的时间序列预测研究提供了一条新的途径.更多还原

【Abstract】 Propose hybrid optimization strategy for the sequence prediction of an area′s human resource data. Currently, neural networks, especially those trained by BP algorithm, are widely used in prediction. However, a feedforward neural network trained by back-propagation(BP) algorithm has a slow study rate and easily gets into the local extremum. Genetic algorithm(GA) is good at finding the global optimum for optimization problems. A hybrid optimization strategy for the sequence prediction is thus proposed. In order to forecast, the back-propagation(BP) algorithm and the genetic algorithm(GA) are combined and applied to training the neural network. The simulation results, compared with those of BP, show that the hybrid method is feasible, thus can provide a new way for the study of strategic planning for human resource.更多还原