Understanding One-Class SVM for Anomaly Detection - AskPython