Research on Sensitivity Analysis of Operational Effectiveness Based on LS-SVM

In efficiency analysis of weapon system, in order to capture and represent the decision maker´s preferences and then to select the most desirable alternative, sensitivity analysis method of operational effectiveness based on LS-SVM is proposed. Firstly, the principle of effectiveness evaluation method based on LS-SVM is discussed. Secondly, to extract learning samples from the MADM problem, an approach to estimate the utility functions for attributes is presented. An optimal selection approach of LS-SVM parameters was put forward based on particle swarm optimization (PSO) algorithm. Finally, the implementation algorithm of sensitivity analysis method of operational effectiveness based on LS-SVM is presented. The proposed method has less number of parameter and is simple and reliable, comparing with SVM. In the end, an example demonstrates the method is feasible and availability.