关键词: 入侵检测, 信息安全, 优化数据处理, 深度学习, 深度信念网络, 未知攻击检测

Abstract: ince those well-known types of intrusions can be detected with higher detection rate at present, but it is very difficult to detect those new-emerging unknown types of network intrusions, a network intrusion detection method based on data optimization deep belief network model was proposed .The above issue will be solved from the method of data processing and model respectively without destroying the existing knowledge and increasing detection time seriously. PMF encoding and MaxMin normalization processing data are applied to DBN model, and the DBN structure is selected through fixing others and changing a parameter, which is based on the benchmark NSL-KDD dataset. Experimental results show that, data optimization impacts on the classification and the intrusion detection method based on DBN algorithm in this paper is self-adaptive. Besides, the detection time of DBN algorithm is similar with that of SVM algorithm and BP neural network model.

Key words: intrusion detection, information security, data optimization, deep learning, deep belief network(DBN), unknown types detection