Research On Situation Prediction Method Based On Joint Hidden Markov And Genetic Algorithm

As the popularity of the Internet has increased year by year,the Internet has become inseparable in all aspects of people's lives.At the same time,cyber attacks are also characterized by gradual complexity,diversification,and scale.The traditional idea of??Relying on a single indicator to evaluate network security performance gradually exposes its limitations.In order to detect and predict the overall security posture of the network,the concept of network security situational awareness has emerged.Cybersecurity situational awareness is a new type of network security theory that comprehensively perceives the entire network situation factor and uses the established theoretical framework and method model to evaluate and predict network risks.The predicted situation value is a reference for the security administrator to make more effective protection measures.In view of the limitations and shortcomings of the current network security situation assessment and prediction methods,this paper proposes an optimization study of the situation assessment and prediction algorithm adapted to the Tim bass model.Mainly done the following research work:By studying the neural network model in the existing situation assessment method,it is found that the initial parameter setting is inaccurate,which leads to the training result falling into the local extremum,the input layer nodes are too much,the model convergence is slow,and the calculation is complicated.Aiming at this,a situation assessment method based on BP neural network combined with bacterial foraging algorithm was proposed.Firstly,the attributes of the input layer are preprocessed to obtain the theoretically acceptable number of nodes,which solves the problem that the calculation of the input layer nodes is too complicated.Then the bacterial foraging algorithm is used to optimize the initial generated neural network parameters.At the beginning of the model training,a relatively better parameter value can be obtained,which solves the problem that the initial parameter setting is improperly caused to cause slow convergence.Finally,the gradient descent method is used to train to obtain an effective model for the final safety situation assessment.2.In the situation prediction method,the existing hidden Markov training model lacks theoretical guidance for the initial parameter setting,and it is easy to make its training result fall into local optimum.Aiming at this,a situation prediction method based on joint hidden Markov model and genetic algorithm is proposed.Firstly,the genetic algorithm is used to optimize the initial value of the hidden Markov model.Then,the Baum Welch algorithm is used to further optimize the model parameters to obtain the hidden Markov model under the maximum likelihood estimation.Finally,the Viter The security situation is predicted by combining the observed values ??with the algorithm.Experiments show that this method can significantly improve the accuracy of network security situation prediction.3.Based on the proposed situation prediction method,a series of algorithm verification experiments are designed.First,based on the collected data,the experimental experiment platform was built using the built experimental platform.Then,the proposed method is used to evaluate and predict the network security situation,and compare it with the traditional situation prediction method.Finally,the effectiveness of the proposed method is verified by predicting the absolute error and other indicators through the network security situation.