Journal of Systems Engineering and Electronics ›› 2020, Vol. 31 ›› Issue (6): 1087-1096.doi: 10.23919/JSEE.2020.000081

• ELECTRONICS TECHNOLOGY •     Next Articles

Reconstruction of time series with missing value using 2D representation-based denoising autoencoder

Huamin TAO(), Qiuqun DENG*(), Shanzhu XIAO()   

  1. 1 National Key Laboratory of Science and Technology on Automatic Target Recognition, College of Electronic Science, National University of Defense Technology, Changsha 410073, China
  • Received:2020-03-15 Online:2020-12-18 Published:2020-12-29
  • Contact: Qiuqun DENG E-mail:Taohmpeach@163.com;dengqiuqun75@163.com;mountbamboo@vip.163.com
  • About author:|TAO Huamin was born in 1972. She received her M.S. degree from National University of Defense Technology, Changsha, China, in 1997. She is currently a researcher at the National Key Laboratory of Science and Technology on Automatic Target Recognition, National University of Defense Technology. Her research interests include signal processing and real-time system design.E-mail: Taohmpeach@163.com||DENG Qiuqun was born in 1991. She received her Ph.D. degree in College of Electronic Science from National University of Defense Technology, Changsha, China, in 2019. She is currently a lecturer at the National Key Laboratory of Science and Technology on Automatic Target Recognition, National University of Defense Technology. Her research interests include optical guidance, target recognition and machine learning. E-mail: dengqiuqun75@163.com||XIAO Shanzhu was born in 1978. He received his M.S. degree from National University of Defense Technology, Changsha, China, in 2002. He is currently a researcher at the National Key Laboratory of Science and Technology on Automatic Target Recognition, National University of Defense Technology. His research interests include automatic target recognition, integrated circuit and real time system. E-mail: mountbamboo@vip.163.com

Abstract:

Time series analysis is a key technology for medical diagnosis, weather forecasting and financial prediction systems. However, missing data frequently occur during data recording, posing a great challenge to data mining tasks. In this study, we propose a novel time series data representation-based denoising autoencoder (DAE) for the reconstruction of missing values. Two data representation methods, namely, recurrence plot (RP) and Gramian angular field (GAF), are used to transform the raw time series to a 2D matrix for establishing the temporal correlations between different time intervals and extracting the structural patterns from the time series. Then an improved DAE is proposed to reconstruct the missing values from the 2D representation of time series. A comprehensive comparison is conducted amongst the different representations on standard datasets. Results show that the 2D representations have a lower reconstruction error than the raw time series, and the RP representation provides the best outcome. This work provides useful insights into the better reconstruction of missing values in time series analysis to considerably improve the reliability of time-varying system.

Key words: time series, missing value, 2D representation, denoising autoencoder (DAE), reconstruction