Context Distribution

Context distribution, the probability distribution governing the contextual information influencing decision-making processes, is a crucial element in various machine learning problems, particularly in contextual bandits and reinforcement learning. Current research focuses on developing algorithms that efficiently learn and adapt to unknown or complex context distributions, often employing techniques like nearest neighbor methods, energy-based models, and kernel density estimation, to achieve optimal regret bounds or accurate best-arm identification. These advancements are significant for improving the performance of sequential decision-making systems in diverse applications, including online advertising, personalized recommendations, and robotics, where contextual information is often crucial but its distribution may be unknown or difficult to characterize.

Papers

October 5, 2024

High Probability Bound for Cross-Learning Contextual Bandits with Unknown Context Distributions
Ruiyuan Huang, Zengfeng Huang
Contextual Bandit Optimal Regret High Probability Bound Context Distribution Cross Inferential Network

August 19, 2024

Contextual Bandits for Unbounded Context Distributions
Puning Zhao, Jiafei Wu, Zhe Liu, Huiwen Wu
Contextual Bandit Optimal Regret Contextual Bandit Problem Minimax Optimal Regret Context Distribution

June 18, 2024

In-Context Learning of Energy Functions
Rylan Schaeffer, Mikail Khona, Sanmi Koyejo
Context Learning Energy Function Context Distribution

March 11, 2024

Acquiring Diverse Skills using Curriculum Reinforcement Learning with Mixture of Experts
Onur Celik, Aleksandar Taranovic, Gerhard Neumann
Reinforcement Learning Mixture Component Expert Knowledge Curriculum Reinforcement Learning Diverse Skill Policy Parametrization Context Distribution

January 21, 2024

Distributed Multi-Task Learning for Stochastic Bandits with Context Distribution and Stage-wise Constraints
Jiabin Lin, Shana Moothedath
Multi Task Learning Linear Bandit Linear Contextual Bandit Heterogeneous Agent Stochastic Bandit State Constraint Context Distribution

January 3, 2024

Optimal cross-learning for contextual bandits with unknown context distributions
Jon Schneider, Julian Zimmert
Contextual Bandit Regret Bound Contextual Bandit Algorithm Price Auction Context Distribution Stochastic Action

December 16, 2023

Stochastic Bayesian Optimization with Unknown Continuous Context Distribution via Kernel Density Estimation
Xiaobin Huang, Lei Song, Ke Xue, Chao Qian
Context Learning Optimization Purpose Bayesian Optimization Kernel Density Context Distribution

September 15, 2022

Best Arm Identification with Contextual Information under a Small Gap
Masahiro Kato, Masaaki Imaizumi, Takuya Ishihara, Toru Kitagawa
Contextual Information Filling Gap Best Arm Identification Adaptive Experiment Context Distribution

June 8, 2022

Learning in Distributed Contextual Linear Bandits Without Sharing the Context
Osama A. Hanna, Lin F. Yang, Christina Fragouli
LeArning Abstract Context Information Linear Contextual Bandit Contextual Linear Bandit Communication Constraint Context Distribution

Context Distribution

Papers

High Probability Bound for Cross-Learning Contextual Bandits with Unknown Context Distributions

Contextual Bandits for Unbounded Context Distributions

In-Context Learning of Energy Functions

Acquiring Diverse Skills using Curriculum Reinforcement Learning with Mixture of Experts

Distributed Multi-Task Learning for Stochastic Bandits with Context Distribution and Stage-wise Constraints

Optimal cross-learning for contextual bandits with unknown context distributions

Stochastic Bayesian Optimization with Unknown Continuous Context Distribution via Kernel Density Estimation

Best Arm Identification with Contextual Information under a Small Gap

Learning in Distributed Contextual Linear Bandits Without Sharing the Context