FreFilterTST: a dynamic channel graph sparsification approach to multivariate time series anomaly detection with frequency-domain restoration
Yi Wang, Jian Jie Zhang, Ming Yang Zhang · 2025
Read FreFilterTST: a dynamic channel graph sparsification approach to multivariate time series anomaly detection with frequency-domain restoration by Yi Wang, Jian Jie Zhang, Ming Yang Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12835106.txt.
Description
Yi Wang, Jian Jie Zhang, and Ming Yang Zhang present a method for spotting anomalies in data streams with many channels at once. Their 2025 paper notes that real problems, from factories to networks, rarely hinge on one variable alone. Existing methods either treat channels separately and miss links, or fuse all channels and swallow noise.
Their framework, FreFilterTST, first rebuilds the input from frequency patches to learn what normal looks like. A transformer with expert sub-networks then filters channels adaptively, thinning the channel graph while keeping fine frequency detail. A frequency-domain restoration step repairs the signal. The design aims to catch subtle faults without drowning in false links.
Catalog record
| CREATOR | Yi Wang, Jian Jie Zhang, Ming Yang Zhang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 102 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |