Research map / Bandits and online learning / Online convex optimization and games
Prediction with expert advice under additive noise
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Studies prediction with expert advice when the loss feedback observed by the decision-maker is corrupted by additive noise, deriving fundamental regret limits.
Problem
Classical prediction with expert advice assumes exact feedback, but real-world feedback is subject to noise, errors, or communication constraints; the paper seeks fundamental performance limits (regret) under such corrupted feedback.
Approach
Analyzes expert-advice online learning where observed losses are perturbed by additive noise, establishing regret limits for this noisy-feedback setting rather than the exact-feedback case.
Experimental setup
Metrics: regret
Builds on: prediction with expert advice
Closest papers by meaning
- A Regret-Variance Trade-Off in Online Learning (NeurIPS 2022 · similarity 0.89)
- Prediction with Corrupted Expert Advice (NeurIPS 2020 · similarity 0.89)
- When Lower-Order Terms Dominate: Adaptive Expert Algorithms for Heavy-Tailed Losses (NeurIPS 2025 · similarity 0.88)
- A second order regret bound for NormalHedge (NeurIPS 2026 · similarity 0.88)
- From Average Sensitivity to Small-Loss Regret Bounds under Random-Order Model (NeurIPS 2026 · similarity 0.87)
- Managing Self-Learning Experts under Per-Round Budget Constraints (NeurIPS 2026 · similarity 0.87)
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.