Zzong's Notes

Home

❯

papers

❯

rl

❯

Offline Reinforcement Learning - Tutorial, Review, and Perspectives on Open Problems

Offline Reinforcement Learning - Tutorial, Review, and Perspectives on Open Problems

2026년 8월 28일1 min read

  • paper link: https://arxiv.org/abs/2005.01643

링크된 언급

1
Pessimistic Reward Models for Off-Policy Learning in Recommendation

Offline Reinforcement Learning - Tutorial, Review, and Perspectives on Open Problems The Optimizer’s Curse - Skepticism and Postdecision Surprise in Dec...

함께 보면 좋은 글

Pessimistic Reward Models for Off-Policy Learning in Recommendation

Introduction 추천 시스템에서 발생하는 loop 는 bias 를 일으킴 (1) collect data, (2) train model, (3) deploy model loop off-policy learning 에서 biased data 로 부터 unbiased learning 을 지향하는 방법에는...

Deep reinforcement learning for search, recommendation, and online advertising - a survey

Paper link arxiv.org/abs/1812.07127 .

A Survey on Reinforcement Learning for Recommender Systems

References paper link: arxiv.org/pdf/2109.10665.pdf Reinforcement Learning based Recommender Systems: A Survey .

Recommendations with Negative Feedback via Pairwise Deep Reinforcement Learning

Discussion Q-learning based offline 학습 방식이고, 모델도 무겁고.

DRN - A Deep Reinforcement Learning Framework for News Recommendation

Abstract 뉴스 추천을 위한 딥러닝 기반의 강화 학습 프레임워크를 제안한다. news feature 들과 user 의 preferences 의 변동성 (dynamic) 을 설명하는 것은 상당히 어렵다.

Reinforcement Learning for Slate-based Recommender Systems - A Tractable Decomposition and Practical Methodology

Empirical Evaluation: Live Experiments YouTube 에 SARSA-TS 알고리즘을 실험 candidate -> ranker 를 거치게 되는데, ranker 의 scoring 함수에서 사용하는 myopic(근시안적) engagement 측정값을 LTV estimate 로 변경함...

Solving Continual Combinatorial Selection via Deep Reinforcement Learning

Abstract S-MDP 문제를 다룬 paper 이다.

Deep Exploration via Bootstrapped DQN

Abstract 해결하려는 문제: 강화학습에서의 효율적인 exploration Randomized value functions offer a promising approach to efficient exploration with generalization, but existing algorithms are not...

Exploring compact reinforcement-learning representations with linear regression

paper Link: arxiv.org/pdf/1205.2606.pdf KWIK Linear Regression KWIK (Knows What It Knows) is a framework for studying supervised learning algorithms and was designed to unify...

Exploration by Random Network Distillation

RL methods work by maximizing the expected return of a policy.