Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Doubly Outlier-Robust Online Infinite Hidden Markov Model
Horace Yiu, Leandro Sánchez-Betancourt, Alvaro Cartea, Gerardo Duran-Martin
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 state-space batching with robust updates | Vanilla HDP-iHMM · Synthetic regression | one-step-ahead RMSE: 101.7 -> 46.1 | Improved |
| add state-space batching on top of WoLF observation robustness | WoLF iHMM · Synthetic regression | one-step-ahead RMSE: 103.8 -> 46.1 | Improved |
| add batching | Vanilla HDP-iHMM · Electricity | one-step-ahead RMSE: 0.57 -> 0.47 | Improved |
| add batching | WoLF iHMM · Electricity | one-step-ahead RMSE: 0.63 -> 0.47 | Improved |
| add batching (tuned B=1 recovers WoLF) | WoLF iHMM · OFI (MSFT) | one-step-ahead RMSE: 0.623 -> 0.616 | No clear effect |
| add batching | Vanilla HDP-iHMM · Synthetic regime detection | PPV: 0.551 -> 0.963 | Improved |
| add batching | Vanilla HDP-iHMM · Synthetic regime detection | TPR: 0.363 -> 0.996 | Improved |
| add batching | Vanilla HDP-iHMM · Synthetic regime detection | detection delay: 0.2 -> 3.4 | Hurt |
| use student-t posterior predictive with unknown variance instead of robust state update | Online iHMM · Well-log | state creation | No clear effect |
| use larger batch size B (ablation) | Batched robust iHMM · OFI (MSFT) | RMSE | Hurt |
| use larger batch size B (ablation) | Batched robust iHMM · Synthetic regime detection | detection delay | Hurt |
Closest papers by meaning
- Outlier-robust Kalman Filtering through Generalised Bayes (ICML 2024 · similarity 0.85)
- Online Bernstein-von Mises theorem (NeurIPS 2026 · similarity 0.84)
- Online MDP with Prototypes Information: A Robust Adaptive Approach (AAAI 2025 · similarity 0.84)
- Learning the Context of Errors: Black-Box Online Adaptation of Time Series Foundation Models (NeurIPS 2026 · similarity 0.83)
- Score-based generative models are provably robust: an uncertainty quantification perspective (NeurIPS 2024 · similarity 0.83)
- Stabilizing Linear Passive-Aggressive Online Learning with Weighted Reservoir Sampling (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.
