Research and Application of BP Algorithm Based on Genetic Algorithm in System Performance Bottleneck Analysis

[摘要] System performance bottleneck analysis is an essential issue in application performance management. With the development of the neural network, BP neural network has been applied to analyze the performance bottleneck of the system, but because BP algorithm is easy to fall into local optimum, the results obtained by this algorithm lack certain accuracy. In the system performance bottleneck, response time and performance counter can effectively help to analyze the performance bottleneck. In this paper, the server performance counter collected related data as sample data, and the improved BP algorithm based on adaptive genetic algorithm is used to model. This method can solve the shortcomings of traditional BP algorithm very well. The experimental results show that the improved BP algorithm based on genetic algorithm is superior to the conventional BP algorithm in the accuracy of system performance bottleneck analysis, and the learning process takes less time on large datasets.

[发布机构] Institute of Network Technology, Engineering Research Center of Information Network, Beijing University of Posts and Telecommunications, Ministry of Education, Beijing, China^1;Institute of Network Technology, Beijing University of Posts and Telecommunications, Beijing, China^2

[效力级别] 材料科学

[关键词] Adaptive genetic algorithms;Application performance;BP neural networks;Improved BP algorithms;Performance bottlenecks;Performance counters;Research and application;Server performance