Anomaly detection method and device

摘要: The invention discloses an anomaly detection method and device. According to the invention, comprehensive log data are comprehensively obtained; features extracted from acquired data; thefeature is processed, model trained generated through SVM algorithm based on processed feature; abnormal behavior automatically detected model; meanwhile, result used for further training a more accurate finally, behaviors detected. Therefore, network data, carried out after continuous evolution learning, sothat accurately detected, problem that cannot be effectively in prior art solved.