Empirical mode decomposition with multivariable time series feature learning for QoS prediction

Yuqi Zhao, Bing Li, Jian Wang, Xiuqing Chen, Yiming Xiong, Zhen Zhang

Abstract: In the past decade, predicting quality of service (QoS) for Web services has been an essential topic in services computing. However, existing QoS prediction approaches typically treat each QoS record as a basic unit without taking the intrinsic properties of the time-series QoS records into account. In fact, a QoS time series is not a smooth distribution but rather a hybrid sequence with multiple frequency features. As a result, we propose AE-mLSTM, a hybrid QoS forecasting method that combines the empirical mode decomposition (EMD) model and the multivariable LSTM model. In addition, AE-mLSTM employs an attention mechanism for multi-task learning, which can learn the shared representation of different tasks and jointly predict the tasks of QoS and timing direction. Experiments conducted on two real-world datasets demonstrate that our approach outperforms several state-of-the-art QoS prediction methods.