Risk Prediction Model Based on Improved AdaBoost Method for Cloud Users

ABSTRACT

Considering the problem how to protect the cloud services from being destroyed by cloud users, the riskprediction model based on improved AdaBoost method is proposed. The risk prediction is regarded as two-class classification problem, and the risk of new cloud users could be predicted by the attributes of historical cloud users. In order to improve the result of predicted, AdaBoost method is adopted in this paper. The error rate of the last training is used to adjust the sample distribution of the next training, which can make the next training have stronger ability of identification for the error-classified samples. At the same time the weight of each weak classifier is set. After all, the strong classifier is generated by combined the weak classifiers through voting, which can improve the overall result of classification. Considering the wrongly-predicted cost, AdaBoost method is improved. The method of cost-sensitive is added into the model in order to minimal the misclassified-cost. Experiments show that the cost-sensitive AdaBoost method has better classification result than the traditional ones and it can predict the risk of the new cloud user effectively and protect the security of the cloud services.

Keywords: