摘要 获取资源与服务性能的关系模型是在云环境中为服务合理分配虚拟资源的关键.然而，训练数据的规模往往显著影响这种非线性关系模型的准确率.针对现有方法不足，提出了将协同过滤推荐(CFR)和支持向量回归(SVR)相结合的服务性能动态建模方法(CSDM).该方法在服务部署与运行时同时训练两种模型，并选择二者中MAE占优的性能模型预测给定资源状态下的服务性能，从而保证预测精度.同时，CSDM引入择优阈值以降低模型训练代价.实验表明，CSDM在不同规模的训练数据上均有较高的预测准确率，且择优阈值对预测精度和建模效率具有显著影响.Abstract：The relationship model between resources and service performance is a key to the proper virtual resource allocation for services in cloud environment. However， the accuracy of these non-linear relationship models is usually significantly influenced by the scale of training data. Aiming at the shortcomings of related work， a dynamic service performance modeling method named CSDM， which combines collaborative filtering recommendation and support vector regression， was proposed. In CSDM， for better accuracy， both performance models were trained at service deployment time and runtime， and the one with lower MAE was selected to estimate the performance under given resource status. In addition， a merit-based threshold was introduced to reduce training costs of performance models. The experimental results showed that CSDM had higher accuracy on different scales of training data， and the merit-based threshold had a significant effect on the prediction accuracy as well as the modeling efficiency.
References
[1]
Lloyd W，Pallickara S，David O，et al.Service isolation vs.consolidation:implications for iaas cloud application deployment［C］//IEEE International Conference on Cloud Engineering.San Francisco，2013:21-30.
[2]
Dejun J，Pierre G，Chi C H.Autonomous resource provisioning for multi-service web applications［C］//Proceedings of the 19th International World Wide Web Conference.New York，2010:471-480.
[3]
Kundu S，Rangaswami R，Dutta K，et al.Application performance modeling in a virtualized environment［C］//IEEE 16th International Symposium on High Performance Computer Architecture.Bangalore，2010:1-10.
[4]
Rao J，Wei Y D，Gong J Y，et al.QoS guarantees and service differentiation for dynamic cloud applications［J］.IEEE Transactions on Network and Service Management，2013，10(1):43-55.
[5]
Zheng Z B，Ma H，Irwin K，et al.QoS-aware web service recommendation by collaborative filtering［J］.IEEE Transactions on Services Computing，2011，4(2):140-152.
[6]
Drucker H，Burges C，Kaufman L，et al.Support vector regression machines［C］//Advances in Neural Information Processing System 9.Cambridge，1997:155-161.
[7]
王宏宇，糜仲春，梁晓艳，等.一种基于支持向量机回归的推荐算法［J］.中国科学院研究生院学报，2007，24(6):742-748.(Wang Hong-yu，Mi Zhong-chun，Liang Xiao-yan，et al.A recommendation algorithm based on support vector regression［J］.Journal of University of Chinese Academy of Science，2007，24(6):742-748.)
[8]
Lorenzi L，Mercier G，Melgani F.Support vector regression with kernel combination for missing data reconstruction［J］.IEEE on Geoscience and Remote Sensing Letters，2012，10(2):367-372.
[9]
Zhou Q，Zhai Y J，Han P.Sequential minimal optimization algorithm applied in short-term load forecasting［C］//IEEE International Conference on Machine Learning and Cybernetics.Hong Kong，2007:2479-2483.
