Scholarlink

To detect anomalies with precision and without prior knowledge in time series, is it better to build a detector from the initial temporal representation or to compute a new, tabular representation using an existing automatic variable construction library? This article addresses this question by conducting an in-depth experimental study for two popular detectors: Isolation Forest and Local Outlier Factor. The results, obtained from experiments on five different datasets, show that the new representation, computed using the tsfresh library, allows Isolation Forest to improve its performance significantly.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Unsupervised Feature Construction for Time Series Anomaly Detection - An Evaluation

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Marine Hamon,
  • Vincent Lemaire,
  • Nour Eddine Yassine Nair-Benrekia,
  • Samuel Berlemont,
  • Julien Cumin

摘要

To detect anomalies with precision and without prior knowledge in time series, is it better to build a detector from the initial temporal representation or to compute a new, tabular representation using an existing automatic variable construction library? This article addresses this question by conducting an in-depth experimental study for two popular detectors: Isolation Forest and Local Outlier Factor. The results, obtained from experiments on five different datasets, show that the new representation, computed using the tsfresh library, allows Isolation Forest to improve its performance significantly.