Structural Classification of Locally Stationary Time Series Based on Second-order Characteristics
Chen Qian, Xiucai Ding, Lexin Li · 2025
Read Structural Classification of Locally Stationary Time Series Based on Second-order Characteristics by Chen Qian, Xiucai Ding, Lexin Li (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/arxiv-2507-04237.txt.
Description
Chen Qian, Xiucai Ding, and Lexin Li sort time series whose character drifts slowly over time, in a 2025 paper. Many signals in science and engineering change their local behavior as they run, which breaks standard classifiers. Their method compares two classes of such series using only second-order traits, the patterns of variation over time. It leans on an autoregressive sketch of each series, then blends many sketches and applies a distance cutoff.
The method needs no minimum training size. In theory, its error rate falls to zero when the classes differ even mildly in those traits. In heavy simulations it beat wavelet, tree, and convolution methods as well as modern deep learning. A real test on brain-wave data for epilepsy classification backed the simulations. The authors call the method fast, practical, and rigorous.
Catalog record
| CREATOR | Chen Qian, Xiucai Ding, Lexin Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 84 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-04237