[2I4-GS-2] Machine learning: Anomaly detection

[2I4-GS-2-01] Development of an Abnormality Predictor Detection Method Using Partial Observation Markov Decision Process (POMDP)

〇Tomoaki Kimura¹, Dinesh Malla^1,2, Masaru Sogabe², katshuyoshi Sakamoto¹, koichi Yamaguchi¹, Tomah Sogabe^1,2 (1. UEC, 2. Grid Inc. )

Keywords:POMDP, abnormality detection

Time series anomaly detection methods are applied for various fields. These methods basically assume distributions for data and users need to set threshold to detect anomality. Otherwise in reinforcement learning, an agent can learn desirable action through interaction with environment and the agent don’t need to know environment. By applying reinforcement learning for anomaly detection, it is possible to detect anomality from trial and error without assumptions. In this paper to deal with time series, we performed anomaly detection using Partially Observable MDP(POMDP). Furthermore, we compared accuracy by changing LSTM steps.