Research map / Bandits and online learning / Contextual bandits and Thompson sampling
True Impact of Cascade Length in Contextual Cascading Bandits
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.
Revisits contextual cascading bandits to clarify how cascade length K affects regret, challenging prior claims about its dependence.
Problem
In contextual cascading bandits, a learner recommends an ordered list (cascade) of items and the user stops at the first attractive one; the role of cascade length K in regret has been unclear.
Approach
Revisits the regret analysis of contextual cascading bandits and argues that, contrary to prior results, the impact of cascade length on regret is different from what was previously claimed. The abstract is truncated, so the exact new bound is not stated.
Surprising findings
- Contrary to prior results that regret grows with cascade length K, the paper reports a different role for K (full abstract not available).
Experimental setup
Metrics: regret
Builds on: cascading bandits
Closest papers by meaning
- Cascading Reinforcement Learning (ICLR 2024 · similarity 0.90)
- Active Context Selection Improves Simple Regret in Contextual Bandits (NeurIPS 2026 · similarity 0.89)
- How Does Variance Shape the Regret in Contextual Bandits? (NeurIPS 2024 · similarity 0.88)
- No-Regret is not enough! Bandits with General Constraints through Adaptive Regret Minimization (ICML 2025 · similarity 0.88)
- Variance-aware Regret Bounds for Stochastic Contextual Dueling Bandits (ICLR 2024 · similarity 0.88)
- Contextual Bandits for Unbounded Context Distributions (ICML 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.