This paper presents a detection algorithm using normalized mutual information feature selection and cooperative evolution of multiple operators based on adaptive parallel quantum genetic algorithm (NMIFS MOP- AQGA). The proposed algorithm is to address the problems that the intrusion detection system (IDS) has lower the detection speed, less adaptability and lower detection accuracy. In order to achieve an effective reduction for high-dimensional feature data, the NMIFS method is used to select the best feature combination. The best features are sent to the MOP- AQGA classifier for learning and training, and the intrusion detectors are obtained. The data are fed into the detection algorithm to ultimately generate accurate detection results. The experimental results on real abnormal data demonstrate that the NMIFS MOP- AQGA method has higher detection accuracy, lower false negative rate and higher adaptive performance than the existing detection methods, especially for small samples sets.

The intrusion detection system (IDS) has lower speed, less adaptability and lower detection accuracy especially for small samples sets. This paper presents a detection model based on normalized mutual antibodies information feature selection and adaptive quantum artificial immune with cooperative evolution of multiple operators (NMAIFS MOP-AQAI). First, for a high intrusion speed, the NMAIFS is used to achieve an effective reduction for high-dimensional features. Then, the best feature vectors are sent to the MOP-AQAI classifier, in which, vaccination strategy, the quantum computing, and cooperative evolution of multiple operators are adopted to generate excellent detectors. Lastly, the data is fed into NMAIFS MOP-AQAI and ultimately generates accurate detection results. The experimental results on real abnormal data demonstrate that the NMAIFS MOP-AQAI has higher detection accuracy, lower false negative rate and a higher adaptive performance than the existing anomaly detection methods, especially for small samples sets.