逐次確率比検定による時系列の構造変化点検出法 : Chow Test との早期検出性の比較 [in Japanese] Structural Change Point Detection Method of Time Series Using Sequential Probability Ratio Test : Comparison with Chow Test in the Ability of Early Detection [in Japanese]
Access this Article
Search this Article
The problem of time series can be classified into three types, in a practical sense. The first problem is how to generate a prediction model that adequately represents the characteristics of the early time series data. The second problem is how to quickly detect the structural change of the time series, where the estimated prediction model does not meet the real data any longer. The third problem is how to correct the time series model after the change detection. This paper focuses on the second problem and proposes a novel method for quick detection of the structural change point in time series. The proposed method is based on a sequential probability ratio test that has been mainly used in the field of quality control. This paper discusses the features of the method from numerical experimentation results. And also, this paper shows its effectiveness, in comparison with the well-known Chow Test.
- IEEJ Transactions on Electronics, Information and Systems
IEEJ Transactions on Electronics, Information and Systems 128(4), 583-592, 2008-04-01
The Institute of Electrical Engineers of Japan