Research map / Time series / Causal and generative time series
Sparse Deep Learning for Time Series Data: Theory and Applications
Mingxuan Zhang, Yan Sun, Faming Liang
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 |
|---|---|---|---|
| PA-RNN (sparse RNN) instead of CF-RNN conformal baseline | PA-RNN · MIMIC-III | prediction interval length: 3.01 -> 2.35 | Improved |
| PA-RNN instead of CF-RNN | PA-RNN · EEG | prediction interval length: 61.86 -> 51.02 | Improved |
| PA-RNN instead of CF-RNN | PA-RNN · COVID-19 | prediction interval length: 733.95 -> 444.93 | Improved |
| PA-RNN instead of CF-RNN | PA-RNN · COVID-19 | joint coverage: 89.7% -> 90.5% | No clear effect |
| PA-RNN instead of MQ-RNN | PA-RNN · EEG | joint coverage: 48.0% -> 94.69% | Improved |
| PA-RNN instead of DP-RNN | PA-RNN · MIMIC-III | joint coverage: 1.2% -> 90.8% | Improved |
| window size 15 (>= true order 7) instead of 1 (ablation) | PA-RNN · NLAR process | hidden-state links: 357 -> 0 | Improved |
| window size 15 instead of 1 (ablation) | PA-RNN · NLAR process | MSPE: 1.056 -> 1.017 | Improved |
| window size 15 instead of 1 (ablation) | PA-RNN · NLAR process | false selection rate: 0 -> 0.23 | Hurt |
| vary window size 1 to 15 (ablation) | PA-RNN · Exponential AR process | MSPE: 1.004 -> 1.001 | No clear effect |
| offline training on 2016-2019 only instead of online retraining of CP methods | PA · French electricity prices | prediction interval length | Improved |
| ACI with different gamma | ACI · French electricity prices | coverage | Mixed |
Closest papers by meaning
- Theoretical analysis of deep neural networks for temporally dependent observations (NeurIPS 2022 · similarity 0.92)
- Interpretable Sparse System Identification: Beyond Recent Deep Learning Techniques on Time-Series Prediction (ICLR 2024 · similarity 0.89)
- Dynamic Sparse Network for Time Series Classification: Learning What to “See” (NeurIPS 2022 · similarity 0.88)
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.87)
- STanHop: Sparse Tandem Hopfield Model for Memory-Enhanced Time Series Prediction (ICLR 2024 · similarity 0.86)
- Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts (ICLR 2025 · similarity 0.86)
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.
