Abnormity detection method and system based on data mining

• Main Authors: CHEN SHAOCHENG, HAI HAIHONG
• Format: Patent
• Language: Chinese; English
• Published: 03.06.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC COMMUNICATION TECHNIQUE; ELECTRIC DIGITAL DATA PROCESSING; ELECTRICITY; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS; TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION

The invention discloses an anomaly detection method and system based on data mining, and the method comprises the steps: collecting the historical behavior data of a client network user, and carrying out the preprocessing; a multi-GRU neural network model is constructed, and hyper-parameters of the multi-GRU neural network model are optimized; training the multi-GRU neural network model through the preprocessed historical behavior data, determining a weight matrix in the multi-GRU neural network, and obtaining an anomaly detection model; and when an interaction request of a client is received, performing anomaly detection on the interaction request of the client through the anomaly detection model. According to the method, the dimension reduction processing is performed on the historical behavior data of the network users through the linear discrimination method, the key behavior characteristics of the network users are reserved while the redundancy between the data is reduced, the multiple GRU neural network