Application of Least Squares Support Vector Regression Trained by Genetic Algorithm in Reliability Prediction of LAN/WLAN Integration Network

Least squares support vector regression trained by genetic algorithm is proposed to predict the reliability of LAN/WLAN integration network,and genetic algorithm is adopted to optimize the parameters of least squares support vector regression in the paper. The influencing factors of network reliability usually include the number of node,the number of link,time delay and reliability of link. The comparison results of the prediction values between LSSVR and RBFNN and the comparison results of the prediction error between LSSVR and RBFNN are given in the paper. lt is indicated that LSSVR has more excellent prediction ability than RBFNN.