Research map / Time series / Causal and generative time series
High-dimensional multivariate forecasting with low-rank Gaussian Copula Processes
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use copula marginal transform instead of mean-scaling (ablation) | GP (low-rank) · average of six datasets | CRPS-Sum ratio: 3.4 -> 1.0 | Helped |
| use copula marginal transform instead of no transformation (ablation) | GP (low-rank) · average of six datasets | CRPS-Sum ratio: 9.5 -> 1.0 | Helped |
| use copula instead of mean-scaling (ablation) | GP (low-rank) · Electricity | CRPS-Sum: 0.022+/-0.000 -> 0.024+/-0.002 | No clear effect |
| use copula instead of mean-scaling (ablation) | GP (low-rank) · Wikipedia | CRPS-Sum: 1.483+/-1.034 -> 0.086+/-0.004 | Helped |
| unroll LSTM per series with shared parameters (GP) instead of single vector LSTM, both with copula and low-rank (ablation) | Vec-LSTM-lowrank-Copula · average of six datasets | CRPS-Sum ratio: 1.7 -> 1.0 | Helped |
| GP parametrization instead of Vec-LSTM (ablation) | low-rank copula model · Traffic | number of parameters: 1.1M -> 44K | Helped |
| use low-rank instead of full-rank output distribution with mean-scaling (ablation) | Vec-LSTM · average of six datasets | CRPS-Sum ratio: 37.6 -> 1.7 | Helped |
Closest papers by meaning
- Multivariate Probabilistic Time Series Forecasting with Correlated Errors (NeurIPS 2024 · similarity 0.87)
- Chimera: Effectively Modeling Multivariate Time Series with 2-Dimensional State Space Models (NeurIPS 2024 · similarity 0.85)
- Structured Matrix Basis for Multivariate Time Series Forecasting with Interpretable Dynamics (NeurIPS 2024 · similarity 0.85)
- TACTiS-2: Better, Faster, Simpler Attentional Copulas for Multivariate Time Series (ICLR 2024 · similarity 0.84)
- SOFTS: Efficient Multivariate Time Series Forecasting with Series-Core Fusion (NeurIPS 2024 · similarity 0.84)
- TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (AAAI 2025 · similarity 0.84)
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.