Online optimization of collaborative web service QoS prediction based on approximate dynamic programming

摘要: More recently, with the increasing demand of web services on World Wide Web used in Internet Things (IoTs), there has been a growing interest study efficient service quality evaluation approaches based prediction strategies to obtain accurate quality-of-service (QoS) values. However, it is obvious that changes significantly under unpredictable network environment. Such impose very challenging obstacles QoS prediction. Most traditional are implemented only using set static model parameters help designer's priori knowledge. Unlike approaches, our algorithm this paper realized by incorporating approximate dynamic programming- (ADP-) online parameter tuning strategy into approach. Through learning and optimization, proposed approach provides automatic capability, prior knowledge or identification not required. Therefore, near-optimal performance can be achieved. Experimental studies carried out demonstrate effectiveness ADP-based