TimeCatcher: A Variational Framework for Volatility-Aware Forecasting of Non-Stationary Time Series DOI: XXXXXXX.XXXXXXX CCS: Computing methodologies Machine learning
Zhiyu Chen, Minhao Liu, Yanru Zhang · 2026
Read TimeCatcher: A Variational Framework for Volatility-Aware Forecasting of Non-Stationary Time Series DOI: XXXXXXX.XXXXXXX CCS: Computing methodologies Machine learning by Zhiyu Chen, Minhao Liu, Yanru Zhang (2026) 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/arxiv-2601-20448.txt.
Description
Zhiyu Chen, Minhao Liu, and Yanru Zhang present TimeCatcher, a forecasting framework for time series that refuse to sit still. Lightweight models based on simple networks forecast well by catching stable trends and seasons. But they rest on an assumption of local calm, so they err on long forecasts of wild series, above all when abrupt swings occur.
Such swings are common in domains like web traffic monitoring. TimeCatcher uses a variational framework that is aware of volatility, so it can brace for unstable stretches. Forecasts that admit when the ground is shaky serve planners better than calm lines through stormy data.
Catalog record
| CREATOR | Zhiyu Chen, Minhao Liu, Yanru Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-20448