earticle

영어

In the cloud computing environment, according to the predicting average response time of service, it can adjust to the follow-up system, so that the response time of the system is acceptable. The traditional methods of predicting average response time of serve mainly include the method of gray predicting and neural network model, but the two methods face several problems, such as longer processing time and unsuitable to larger volatility data. According to the above problems, the paper proposes the method of predicting average response time of cloud service based on the MGM (1, N) - BP neural network, the combination of two methods of predicting can use less sample information, it can get a high precision of predicting result and it can also predict the volatile system. Experimental results show the feasibility and effectiveness of the method.