Research map / Learning theory and algorithms / Combinatorial and search optimization
Reliever: Relieving the Burden of Costly Model Fits for Changepoint Detection
Chengde Qian, Guanghui Wang, Changliang Zou
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Reliever (r=0.9) model-fit reuse | SN · KDE changepoint | density-fit time (centiseconds): 447.1 -> 147.3 | Improved |
| add Reliever (r=0.9) model-fit reuse | WBS · KDE changepoint | detection error: 38.9 -> 39.6 | No clear effect |
| add Reliever (r=0.9) model-fit reuse | DP · High-dim linear changepoint | runtime: 7,700.0 -> 82.0 | Improved |
| add Reliever (r=0.9) model-fit reuse | DP · High-dim linear changepoint | detection error: 12.6 -> 13.4 | No clear effect |
| add Reliever (r=0.9) model-fit reuse | DCDP Stage I-II · High-dim linear changepoint | runtime: 33.4 -> 11.2 | Improved |
| add Reliever (r=0.9) model-fit reuse | OP · Univariate nonparametric Model (A), n=1000 | time (seconds): 2804.17 -> 22.65 | Improved |
| add Reliever (r=0.9) model-fit reuse | PELT · Univariate nonparametric Model (B), n=1000 | time (seconds): 65.85 -> 8.56 | Improved |
| lower coverage ratio from 0.9 to 0.5 (ablation) | Reliever with OP · Univariate nonparametric Model (A), n=1000 | UE: 49.16 -> 69.16 | Hurt |
| lower coverage ratio from 1.0 to 0.9 (ablation) | Reliever with OP · Univariate nonparametric Model (A), n=1000 | time (seconds): 2804.17 -> 22.65 | Improved |
| raise coverage ratio r from 0.5 to 1 (fewer proxy approximations) (ablation) | SN model fits · Number of model fits | model fits: 440 -> 686,206 | Hurt |
| use Reliever (r=0.9) instead of two-step method | single changepoint high-dim linear · High-dim linear single changepoint (n=1200) | detection error: 18.6 -> 9.2 | Improved |
| lower coverage ratio from 0.9 to 0.3 (ablation) | Reliever · High-dim linear single changepoint (n=1200) | time (10ms): 480.5 -> 64.1 | Improved |
Closest papers by meaning
- Optimistic Search: Change Point Estimation for Large-scale Data via Adaptive Logarithmic Queries (JMLR 2024 · similarity 0.89)
- An Efficient Two Step Algorithm for High Dimensional Change Point Regression Models Without Grid Search (JMLR 2019 · similarity 0.85)
- Divide and Conquer Dynamic Programming: An Almost Linear Time Change Point Detection Methodology in High Dimensions (ICML 2023 · similarity 0.85)
- Bayesian Model Selection for Change Point Detection and Clustering (ICML 2018 · similarity 0.85)
- TSB-SEG: A Systematic Time-Series Segmentation Benchmark (NeurIPS 2026 · similarity 0.85)
- Segment Any Change (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
