Research on weapon system cost forecasting model based on chaos optimization LSSVM

Forecasting weapon system cost accurately has great meaning in determining weapons´ appropriate price, reducing the cost risk and raising the efficiency of equipment expense utilization. The least squares support vector machine (LSSVM) was applied to forecast weapon system cost, and the chaos optimization algorithm, which is regarded as a good optimization method, was used to optimize the penalty function and kernel function of LSSVM in order to resolve the problem of random selection of parameters in LSSVM. Further, the weapon system cost forecasting model based on chaos optimization LSSVM was established. The practical results show that, compared with neural network and support vector machine without optimization, the training and testing accuracy of the proposed model is better. Therefore, the model established above is scientific and valid for weapon system cost forecasting.