Zzong's Notes

방문

❯

❯

❯

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

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

최초 작성 2026-06-14마지막 수정 2026-08-29읽는 데 1분

노트 편집 ↗

Paper link
- https://arxiv.org/abs/1812.07127

함께 보면 좋은 글

A Survey on Reinforcement Learning for Recommender Systems

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

DRN - A Deep Reinforcement Learning Framework for News Recommendation

뉴스 추천을 위한 딥러닝 기반의 강화 학습 프레임워크를 제안한다.

Recommendations with Negative Feedback via Pairwise Deep Reinforcement Learning

Discussion Q-learning based offline 학습 방식이고, 모델도 무겁고.. negative feedback 을 사용했다고 하는데, 관련 연구 내용이 없음 강화학습 주제이지만, MAB 와 비교가 없음 - [ ] MAB 는 MDP 랑 무슨 관계지? Abstract we propose a...

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

YouTube 에 SARSA-TS 알고리즘을 실험 candidate - ranker 를 거치게 되는데, ranker 의 scoring 함수에서 사용하는 myopic(근시안적) engagement 측정값을 LTV estimate 로 변경함 candidate scoring 쪽은 그대로 유지

Solving Continual Combinatorial Selection via Deep Reinforcement Learning

S-MDP 문제를 다룬 paper 이다. S-MDP 란 매 step 마다 item 들의 부분 집합을 선택하는 MDP 문제

Deep Exploration via Bootstrapped DQN

해결하려는 문제: 강화학습에서의 효율적인 exploration

Pessimistic Reward Models for Off-Policy Learning in Recommendation

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 importance...

Exploration by Random Network Distillation

RL methods work by maximizing the expected return of a policy. In reality it is often impractical to engineer dense reward functions for every task one wants an RL agent to solve.

Exploring compact reinforcement-learning representations with linear regression

paper Link: https://arxiv.org/pdf/1205.2606.pdf

Wide & Deep Learning for Recommender Systems

A wide linear model and a deep neural network are trained together to both memorize and generalize a recommendation systems.