Int J Performability Eng ›› 2019, Vol. 15 ›› Issue (11): 2927-2935.doi: 10.23940/ijpe.19.11.p11.29272935

Previous Articles     Next Articles

A Novel Ensemble Forecasting Algorithm based on Distributed Deep Learning Network

Tao Ma*, Fen Wang, Yanshan Tian, Yan Ma, and Xu Ma   

  1. School of Mathematical and Computer Science, Ningxia Normal University, Guyuan, 756000, China
  • Submitted on ; Revised on ; Accepted on
  • Contact: * E-mail address: matao@nxnu.edu.cn
  • About author:Tao Ma received his Ph.D. in computer science from the School of Information Science and Engineering at Lanzhou University in 2017. He is currently an associate professor in the School of Mathematical and Computer Science at Ningxia Normal University. His research interests include data mining and algorithm optimization.Fen Wang received her MSc. degree in computer science and technology from Shaanxi Normal University in 2004 and her M.S. degree in computer graphics theory from Ningxia University in 2010. She is currently an associate professor and senior engineer at Ningxia Normal University. Her research interests include data mining and face recognition algorithms.Yanshan Tian received his Ph.D. in computer science from the School of Information Science and Engineering at Lanzhou University in 2018. He is currently an associate professor in the School of Mathematics and Computer Science at Ningxia Normal University. His research interests include embedded systems and parallel computing.Yan Ma received her Ph.D. in basic physics theory from Shaanxi Normal University in 2018. She is currently an associate professor and senior engineer at Ningxia Normal University. Her research interests include ultrasonic engineering and ultrasonic cavitation.Xu Ma received his master's degree in computer science from Xi'an University of Electronic Technology. He is currently a professor in the School of Mathematics and Computer Science at Ningxia Normal University. His research interests include cloud computing and smart computing.

Abstract: This paper proposes an ensemble model based on distribution deep learning network. The ensemble model is composed of deep belief network (DBN) for reconstructing original data, and the bidirectional long short-term memory (BLSTM) method is used for prediction due to its good results in big data applications. The dynamic weighting strategies are proposed and applied to the sub models of the ensemble by a weighted least square method. The weight update with variable training sets and the predictions for each ensemble are obtained from the distributed computing engine Apache Spark. The performance of the proposed model is evaluated on wind data on the wind farm of the Hexi Corridor in China. The simulation results show that the dynamic ensemble algorithm performs well, which is a very valuable result for the forecasting of big data time series. Furthermore, the results are successfully compared with back propagation neural Network (BPNN), LSTM, BLSTM, and stacked LSTMs with memory between batches (SBLSTM), improving the accuracy of prediction.

Key words: distributed deep learning network, wind time series, ensemble, forecasting, big data