ONLINE PERSONALIZED QOS PREDICTION APPROACH FOR CLOUD SERVICES

Personalized Quality-of-Service (QoS) prediction is an indispensable technique to select suitable services for service-based cloud applications. Considering the dynamic nature of services, efficiently and accurately predicting QoS value becomes an urgent and crucial research issue. In this paper, we propose an online personalized QoS prediction approach for cloud service, namely online learning based matrix factorization (OLMF). We build the objective function of online matrix factorization and use stochastic gradient descent algorithm to solve the function. Extensive experiments are conducted on real world public datasets, which verify the effectiveness and efficiency of our proposed approach.