A Finite Time Analysis of Thompson Sampling for Bayesian Optimization with Preferential Feedback
2026-04-29T07:24:06Z•af7e16ceee196fd560871699a744a46dc97414cd296c94f32e89880c72bf3e53
DFSOSarxivbanditsbayesian-optimizationchange-point-detectiondeflation-freedueling-kerneldynamical-systemsedsvmeffective-dimensionerdos-renyigraph-learningorthogonality-constrained-optimizationphysics-informed-neural-networkspreference-learningregret-boundsresidual-analysisside-observation-graphssparse-optimal-scoringspectral-banditsstat-mlsupport-vector-machinessvmthompson-samplingtransformers-analysis','relu-to-softmax','softmax-attention','ne
What happened
Feed of arXiv stat.ML submissions (2026-04-29). Papers cover new algorithms and theoretical results across Bayesian optimization (Thompson Sampling for pairwise/preference feedback using a dueling kernel and finite-time guarantees, double-TS pairing), SVM variants (Elite-Driven SVMs that incorporate reference/slack priors, C-EDSVM and LS-EDSVM), multi-armed bandits with side-observations (Erdős–Rényi observation model with regret bounds), spectral bandits on graphs (effective-dimension analysis for graph-smooth payoffs), physics-informed neural networks for change-point detection (residual‑an/
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_stat_ml
- Record identifier
- af7e16ceee196fd560871699a744a46dc97414cd296c94f32e89880c72bf3e53
- Enrichment time
- 2026-04-29T07:24:06Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.