On Detecting Multiple Simultaneous Change-points in High Dimensional Non-Stationary Time Series
Abstract
This paper studies the detection of multiple simultaneous (systematic) change points for high-dimensional nonstantionary economic and financial time series data. The analytic framework used is based on the standard and adaptive fused group lasso method, where the mixed L_{2,1} penalty is either uniform or re-weighted by data-dependent weights. This paper shows that, under appropriate conditions, this approach is L_2 consistent and, by adopting the data-dependent weights, could correctly select the change points with probability approaching unity (L_0 consis- tency). It quantifies the conditions on the interplay among the averaged minimum magnitude of structural changes, the number of change points and the number of observations for consistently discovering the change points. The performance of this approach is illustrated via an analysis of a large panel of U.S. economic and financial time series data over the past 50 years.
Related
- Simultaneous Change-Point Inference for High-Dimensional Functional Time Series
- Multiple Change Point Detection in the Mean and Covariance of High-Dimensional Matrix Time Series
- SCAN: Sequentially Detecting Change-points via Adaptive Nonparametric Inference
- Smooth stationary water waves with multiple vortex spikes
- Optimal detection of general moment changes: Simultaneous mean and covariance change detection and beyond
BibTeX
@misc{song2026detecting,
title = {{On Detecting Multiple Simultaneous Change-points in High Dimensional Non-Stationary Time Series}},
author = {Richard Song},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.15479},
eprint = {2609.15479},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.15479},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - On Detecting Multiple Simultaneous Change-points in High Dimensional Non-Stationary Time Series
AU - Richard Song
PY - 2026
DA - 2026/09/14
PB - arXiv
DO - 10.48550/arxiv.2609.15479
UR - https://arxiv.org/abs/2609.15479
AB - This paper studies the detection of multiple simultaneous (systematic) change points for high-dimensional nonstantionary economic and financial time series data. The analytic framework used is based on the standard and adaptive fused group lasso method, where the mixed L_{2,1} penalty is either uniform or re-weighted by data-dependent weights. This paper shows that, under appropriate conditions, this approach is L_2 consistent and, by adopting the data-dependent weights, could correctly select the change points with probability approaching unity (L_0 consis- tency). It quantifies the conditions on the interplay among the averaged minimum magnitude of structural changes, the number of change points and the number of observations for consistently discovering the change points. The performance of this approach is illustrated via an analysis of a large panel of U.S. economic and financial time series data over the past 50 years.
ER -