Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Communication Efficient Distributed Learning for Kernelized Contextual 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 |
|---|---|---|---|
| learn a shared model with immediate aggregation (OneKernelUCB) instead of independent per-client models | NKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | cumulative regret | Helped |
| communicate every step (OneKernelUCB) instead of never | NKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | communication cost | Hurt |
| transfer raw data at synchronization (DisKernelUCB) | Approx-DisKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | communication cost | Hurt |
| use linear model (DisLinUCB) instead of kernelized model | DisKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | cumulative regret | Hurt |
| use linear model (DisLinUCB) instead of kernelized model | DisKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | communication cost | Helped |
| replace exact kernel estimators with Nystrom approximation over a shared dictionary (Approx-DisKernelUCB) | DisKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | cumulative regret | No clear effect |
| replace exact kernel estimators with Nystrom approximation over a shared dictionary (Approx-DisKernelUCB) | DisKernelUCB · synthetic, UCI, MovieLens and Yelp datasets | communication cost | Helped |
| use Approx-DisKernelUCB instead of DisLinUCB | DisLinUCB · synthetic, UCI, MovieLens and Yelp datasets | cumulative regret | Helped |
Closest papers by meaning
- Learning Kernelized Contextual Bandits in a Distributed and Asynchronous Environment (ICLR 2023 · similarity 0.95)
- Kernel Methods for Cooperative Multi-Agent Contextual Bandits (ICML 2020 · similarity 0.92)
- Distributed Contextual Linear Bandits with Minimax Optimal Communication Cost (ICML 2023 · similarity 0.91)
- Distributed Multi-Agent Bandits Over Erdős-Rényi Random Networks (NeurIPS 2025 · similarity 0.88)
- Communication-Efficient Collaborative Regret Minimization in Multi-Armed Bandits (AAAI 2024 · similarity 0.87)
- Federated Contextual Cascading Bandits with Asynchronous Communication and Heterogeneous Users (AAAI 2024 · 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.