A support vector machine-based pattern recognizer using selected features for control chart patterns analysis

In this paper we review two implementation modes of control chart pattern recognition and introduce a new research problem concerning pattern displacement problem in the ¿recognition only when necessary¿ mode. A set of features are developed by taking the pattern displacement into account. Simulation studies indicate that an SVM-based pattern recognizer with features as input vector performs significantly better than that of using raw data as inputs.