Research On Neural Network Intrusion Detection Algorithm Based On Distributed Learning

With the rapid development of network, how to ensure the security ofnetwork information becomes more and more important. As an active securitytechnology, intrusion detection has been paid more and more attention to. As akind of important method, applied in intrusion detection has a unique advantage.Through the analysis and training of a large number of instance data, neuralnetwork can learn knowledge by using the method of training. According to theexisting instances, neural network can independently grasp and analyse therelationship between each variable and instance in the system, and do not need toknow details of data distribution and parsing.This paper introduces the function, classification and method of intrusiondetection. In detail, it expounds the concept and principle of neural network, andthe research content of neural network is also elaborated. Then this paperintroduces the principle importantly, steps and processes of BP algorithm, bycomparing the advantages and disadvantages of the algorithm, and combiningwith BP neural network model, to improve the existing algorithms.First of all, the principle of neural network is discussed, and the traditionalBP network learning algorithm and BP neural network adaptive learningalgorithm are researched. Combining the advantages of two algorithms, thedistributed neural network self-learning algorithm is proposed, which is a kind ofintrusion detection algorithm using the method of distributed learning to optimizethe BP neural network algorithm. Using this algorithm to study and test thenetwork intrusion data, it solves the problem that directly using BP learningcaused by the training sample size too large and difficult to convergence. At thesame time, the sample training time is shortened, and the BP neural networkclassification accuracy is improved.Secondly, based on the research of the improved algorithm, this paper gives the specific steps of the algorithm, and uses the improved algorithm to establishmathematical model which is used to analysising and forecasting. Compared withthe traditional BP network learning algorithm and BP neural network adaptivelearning algorithm, verify the effectiveness and feasibility of the improvedalgorithm.Finally, the algorithm is applied to intrusion detection. Through appropriatetest method, use the sample data of this paper adopted to verify the example.Through the results of the testing data, it verifies the performance of thedistributed neural network self-learning algorithm, and comes to the conclusion.