DBN And MDBoost2 Used In Intrusion Detection

Due to the rapid development of the Internet, the scale of the network data is increasing day by day and the network topology is becoming more and more complex, which results in a large number of network data and complex data characteristics. And it also increases the redundancy of the dat. All of these greatly increase the difficulty of intrusion detection. Therefore, how to effectively improve the accuracy of prediction is a major problem of intrusion detection.Nowadays, the model of machine learning and data mining has been widely used in the field of intrusion detection. However, because the network data has a large scale and high dimension, and simple model is difficult to handle this problem, which prone to false positives and false negatives. In recent years, deep learning have a large number of applications in the image recognition and text mining. Deep learning is very powerful in the generalization and abstraction of data features. So it can effectively improve the accuracy of prediction when deep learning is used in intrusion detection. However, the predictive power of a single classifier is limited. If by some integrated algorithm of multiple weak classifiers are combined into a strong classifier, it can effectively improve the prediction accuracy of intrusion detection. Aiming at the challenge of intrusion detection, this paper mainly studies the application of deep learning and integrated algorithm in intrusion detection.In the application of the existing machine learning in intrusion detection, combined with the theory of deep learning, this paper discusses the application of BP neural network and Deep belief network in intrusion detection and compares the algorithms’ validity in intrusion detection with the experimental results. However, it is difficult to accurately predict the complex network data with a single weak classifier. In order to improve the accuracy of model prediction, the paper focuses on the integration algorithm, that is, how to use a single weak classifier to become a strong classifier. Then, an improved algorithm is proposed to improve the MDBoost2 algorithm. At the same time, the paper compares the results of the simulation experiments in the intrusion detection system. Then,an improved algorithm is proposed to improve the MDBoost2 algorithm.At the same time, this paper proposes a new intrusion detection model, which is a hybrid model with a deep confidence network as a weak classifier,and the key modules of the model are analyzed in detail. Finally,the simulation results are compared with the results of the intrusion detection system. It is concluded that the MDBoost2 algorithm model with DBN as the weak classifier has better performance than the same type of model in intrusion detection...