Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Detecting and Adapting to Irregular Distribution Shifts in Bayesian Online Learning
Aodong Li, Alex James Boyd, Padhraic Smyth, Stephan Mandt
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 |
|---|---|---|---|
| Increase temperature beta in VBS (ablation) | VBS (K=1, greedy) · Basketball player tracking | number of detected change points | Improved |
| Use larger beam size in VBS (ablation) | VBS · CIFAR-10 and SVHN with covariate shifts | classification performance | Improved |
| Compare VBS against Bayesian forgetting (BF) | VBS · CIFAR-10 and SVHN with covariate shifts | classification performance | No clear effect |
| Add VBS to online linear regression versus baselines | Online Bayesian linear regression · Malware, SensorDrift, Elec2 | mean cumulative absolute error | Improved |
| Use VBS instead of baselines for trajectory prediction | Bayesian transition matrix · Basketball player tracking | time-averaged predictive log-likelihood | Improved |
| Vary prior log-odds xi_0 to sparsify embedding updates, VBS vs binning oracle (ablation) | Dynamic word embeddings with VBS · Document dating | dating accuracy vs sparsity | Improved |
| Add tempering (VGS) to online Bayesian regression instead of plain VCL (ablation) | Online Bayesian linear regression · 1D regression with a shift | adaptation to shifted function | Improved |
Closest papers by meaning
- Online Learning via Learned Latent Bayesian Tracking (NeurIPS 2026 · similarity 0.87)
- Online Bernstein-von Mises theorem (NeurIPS 2026 · similarity 0.87)
- Quantifying Uncertainty in the Presence of Distribution Shifts (NeurIPS 2025 · similarity 0.86)
- Optimal Online Change Detection via Random Fourier Features (NeurIPS 2025 · similarity 0.85)
- Label Shift Meets Online Learning: Ensuring Consistent Adaptation with Universal Dynamic Regret (CVPR 2025 · similarity 0.85)
- Stationary Latent Weight Inference for Unreliable Observations from Online Test-Time Adaptation (ICML 2024 · similarity 0.85)
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.
