An Effective and Novel Weighted Support Vector Machine Method for Control Chart Pattern Recognition

Control chart pattern recognition is the method to realize quality online monitoring and diagnosis of production process. For the conditions that the number of existing normal mode products is much higher than the abnormal ones during the actual manufacturing process, we proposed a method about WSVM (Weighted Support Vector Machines) for dynamic process of abnormal pattern recognition based on PCA (Principal Component Analysis). We put the proposed method into our experiment, the experimental simulation results show that this method proposed in this paper has a big advantage over the existing SVM (Support Vector Machine) on highly imbalanced classification problem, which suitable for quality monitoring and diagnosis of dynamic production process.