Research map / Time series / Causal and generative time series
Multivariate Bayesian Structural Time Series Model
Jinwen Qiu, S. Rao Jammalamadaka, Ning Ning
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 |
|---|---|---|---|
| use multivariate MBSTS instead of univariate BSTS | Bayesian structural time series · Simulated multi-target series | cumulative absolute one-step-ahead prediction error | Improved |
| use MBSTS instead of ARIMAX/MARIMAX | Time series forecasting · Simulated series with seasonality or cycle | cumulative absolute one-step-ahead prediction error | Improved |
| use MBSTS instead of benchmarks on series with only local or linear trend | Time series forecasting · Simulated series with local/linear trend only | cumulative absolute one-step-ahead prediction error | No clear effect |
| increase correlation among target series (ablation) | MBSTS versus univariate models · Simulated series, correlations 0 to 0.8 | cumulative absolute one-step-ahead prediction error | Improved |
| add more target series (three or four) (ablation) | MBSTS · Simulated series | cumulative absolute one-step-ahead prediction error | Improved |
| deseasonalize predictors with STL (ablation) | MBSTS · Stock portfolio (BOA, COF, JPM, WFC) | cumulative one-step-ahead prediction error | Hurt |
| use MBSTS instead of BSTS, ARIMAX, MARIMAX | Stock max log return forecasting · Stock portfolio (BOA, COF, JPM, WFC) | cumulative one-step-ahead prediction error | Improved |
| increase sample size (ablation) | MBSTS regression coefficient estimation · Simulated data, 50 to 3200 observations | estimation error and standard error | Improved |
Closest papers by meaning
- A Set-Sequence Model for Time Series (NeurIPS 2026 · similarity 0.88)
- Structured Matrix Basis for Multivariate Time Series Forecasting with Interpretable Dynamics (NeurIPS 2024 · similarity 0.85)
- Chimera: Effectively Modeling Multivariate Time Series with 2-Dimensional State Space Models (NeurIPS 2024 · similarity 0.85)
- AdaPTS: Adapting Univariate Foundation Models to Probabilistic Multivariate Time Series Forecasting (ICML 2025 · similarity 0.84)
- Multivariate Probabilistic Time Series Forecasting with Correlated Errors (NeurIPS 2024 · similarity 0.84)
- Latent Variable Estimation in Bayesian Black-Litterman Models (ICML 2025 · 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.
