Research of Security Detection for Networked Manufacturing Based on Optimized Support Vector Machine

Security detections for networked manufacturing can improve availability of security configuration and also lower life cycle cost. But the threat level should be classified scientifically before a safety decision could be given. In this paper, a new machine learning method suitable for small-sample pattern recognition, called least squares support vector machine, is studied, and the optimization method for selecting the parameters of least squares support vector machines is presented. Then we built a security detection multi-class classifier for networked manufacturing and tested the model by classifying new security detections. The simulation results show that this method is efficient and feasible.