並列タイトル等Statistical Machine Learning Approaches to Change Detection
一般注記In this thesis, we focus on unsupervised change detection, and propose two novel approaches in distributional and structural change detection respectively. First, we propose a statistical change-point detection algorithm based on non-parametric divergence estimation between time-series samples from two retrospective segments. Our method uses the relative Pearson divergence as a divergence measure, and it is accurately estimated by a method of density ratio estimation. Second, we propose a method for detecting changes in Markov network structure between two sets of samples. Instead of naively fitting two Markov network models separately to the two data sets and figuring out their difference, we directly learn the network structure change by estimating the ratio of Markov network models. The usefulness of both methods is demonstrated by experiments. The results show that the proposed methods can successfully capture the changes of patterns in many real-world applications.
identifier:oai:t2r2.star.titech.ac.jp:50230928
一次資料へのリンクURLhttp://t2r2.star.titech.ac.jp/rrws/file/CTT100666815/ATD100000413/main.pdf （fulltext）
連携機関・データベース国立情報学研究所 : 学術機関リポジトリデータベース（IRDB）（機関リポジトリ）