Reconstructive reservoir computing for anomaly detection in time-series signals

DOI Web Site 研究データあり 参考文献37件 オープンアクセス
  • Kato Junya
    Department of Electrical Engineering and Information Systems, The University of Tokyo
  • Tanaka Gouhei
    Department of Computer Science, Nagoya Institute of Technology International Research Center for Neurointelligence (IRCN), The University of Tokyo
  • Nakane Ryosho
    Department of Electrical Engineering and Information Systems, The University of Tokyo
  • Hirose Akira
    Department of Electrical Engineering and Information Systems, The University of Tokyo

説明

<p>We propose reconstructive reservoir computing (RRC) which performs better anomaly detection in time-series signals than forecasting-based methods. In this paper, reconstruction means that a neural network generates past input signals. RRC reconstructs a past normal signal for anomaly detection using an echo state network which can learn quickly and stably. We expect that it is easier to restore a past normal signal than to predict an unknown future normal signal. For anomaly detection, we compute an instantaneous reconstruction error. The reconstruction error larger than a threshold is a sign of anomaly. We conduct experiments using sound data obtained from a pump. In the experiments, we pay attention to a time lag between input and output to be reconstructed since we assume that an excessive time lag makes reconstruction difficult due to signal attenuation in the network. Experimental results show that if the time lag is moderate, the reconstruction error of the normal signal is lower than the forecasting error of the same signal. Furthermore, we show that RRC with the appropriate time lag has a better anomaly detection performance index than forecasting-based methods.</p>

収録刊行物

参考文献 (37)*注記

もっと見る

キーワード

詳細情報 詳細情報について

問題の指摘

ページトップへ