Less is More: Efficient Time Series Dataset Condensation via Two-fold Modal Matching
Summary: TimeDC compresses large time-series datasets so models trained on condensed data achieve performance comparable to full datasets. It uses two-fold modal matching—decomposition-driven frequency matching to preserve temporal/frequency structure and curriculum training trajectory matching with an expert-buffer to cut memory/compute during condensation. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Hao Miao (Aalborg University)
- 2. Ziqiao Liu (University of Electronic Science and Technology of China)
- 3. Yan Zhao (Aalborg University)
- 4. Chenjuan Guo (East China Normal University)
- 5. Bin Yang (East China Normal University)
- 6. Kai Zheng (University of Electronic Science and Technology of China)
- 7. Christian S. Jensen (Aalborg University)
BibTeX Citation
@article{miao_vldb25,
title = {{Less is More: Efficient Time Series Dataset Condensation via Two-fold Modal Matching}},
author = {Miao, Hao and Liu, Ziqiao and Zhao, Yan and Guo, Chenjuan and Yang, Bin and Zheng, Kai and Jensen, Christian S.},
journal = {PVLDB},
series = {{VLDB} '25},
volume = {18},
number = {2},
pages = {226--238},
doi = {10.14778/3705829.3705841},
url = {https://doi.org/10.14778/3705829.3705841},
year = {2025}
}Incoming Citations (Sorted by Pagerank)
Showing 3 of 3 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 11,285 | TEAM: Topological Evolution-aware Framework for Traffic Forecasting | 2025 | VLDB | 4.9793485e-05 |
| 13,640 | Fully Automated Correlated Time Series Forecasting in Minutes | 2025 | VLDB | - |
| 13,642 | A Memory Guided Transformer for Time Series Forecasting | 2025 | VLDB | - |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 20 of 20 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 8,286 | Anticipatory DTW for Efficient Similarity Search in Time Series Databases | 2009 | VLDB |
| 2 | 6,627 | Interactive Time Series Exploration Powered by the Marriage of Similarity Distances | 2017 | VLDB |
| 3 | 11,847 | Scalable Time Series Compound Infrastructure | 2022 | SIGMOD |
| 4 | 8,875 | The Inherent Time Complexity and An Efficient Algorithm for Subsequence Matching Problem | 2022 | VLDB |
| 5 | 12,762 | Parsimonious Linear Fingerprinting for Time Series | 2010 | VLDB |
| 6 | 7,754 | Uncertain Time-Series Similarity: Return to the Basics | 2012 | VLDB |
| 7 | 9,077 | A Multi-Scale Decomposition MLP-Mixer for Time Series Analysis | 2024 | VLDB |
| 8 | 11,717 | LightTS: Lightweight Time Series Classification with Adaptive Ensemble Distillation | 2023 | SIGMOD |
| 9 | 6,682 | Hierarchical Residual Encoding for Multiresolution Time Series Compression | 2023 | SIGMOD |
| 10 | 5,102 | Efficient Temporal Pattern Mining in Big Time Series Using Mutual Information | 2022 | VLDB |