Research map / Bandits and online learning / Multi-armed bandits
Improved Confidence Bounds for the Linear Logistic Model and Applications to Bandits
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 the variance-dependent fixed-design confidence bound based on Fisher information | Linear logistic MLE confidence intervals · Fixed-design linear logistic model | confidence-width dependence: O(κ^{-1}‖x‖_{V_t^{-1}}√log(1/δ)) -> O(‖x‖_{H_t(θ*)^{-1}}√log(t/δ)) | Helped |
| Use RAGE-GLM instead of the passive baseline | Transductive logistic pure exploration · Zappos pairwise: open | samples: 2.69e+08 -> 9.17e+07 | Helped |
| Use RAGE-GLM instead of the passive baseline | Transductive logistic pure exploration · Zappos pairwise: comfort | samples: 8.54e+06 -> 4.55e+06 | Helped |
| Use RAGE-GLM instead of RAGE-GLM-R | Transductive logistic pure exploration · Zappos pairwise: comfort | samples: 3.34e+06 -> 4.55e+06 | Hurt |
| Replace the SupCB-GLM leading regret dependence with SupLogistic's confidence-bound design | Contextual logistic bandits · K-armed contextual bandits | leading regret bound: Õ(κ^{-1}√(dT log(K))) -> Õ(√(dT log(K))) | Helped |
| Use the separate Φ bucket to estimate confidence widths | SupLogistic · K-armed contextual bandits | confidence-bound validity | Helped |
| Use a separate bucket for confidence-width estimation rather than estimating widths from the same adaptive bucket | SupLogistic · K-armed contextual bandits | confidence-bound validity | Helped |
Closest papers by meaning
- Improved Online Confidence Bounds for Multinomial Logistic Bandits (ICML 2025 · similarity 0.92)
- Improved Optimistic Algorithms for Logistic Bandits (ICML 2020 · similarity 0.91)
- Finite-Time Logarithmic Bayes Regret Upper Bounds (NeurIPS 2023 · similarity 0.90)
- Bayesian Regret Minimization in Offline Bandits (ICML 2024 · similarity 0.89)
- Noise-Adaptive Confidence Sets for Linear Bandits and Application to Bayesian Optimization (ICML 2024 · similarity 0.89)
- Improved Regret Bounds for Linear Bandits with Heavy-Tailed Rewards (NeurIPS 2025 · similarity 0.88)
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.