50 citations · 108 across the 18 of their papers we have counts for
10 papers · 1 filter
Reinforcement Learning with History-Dependent Dynamic Contexts
Guy Tennenholtz, Nadav Merlis, Lior Shani +2
We introduce Dynamic Contextual Markov Decision Processes (DCMDPs), a novel reinforcement learning framework for history-dependent environments that generalizes the contextual MDP…
Towards Content Provider Aware Recommender Systems: A Simulation Study on the Interplay between User and Provider Utilities
Ruohan Zhan, Konstantina Christakopoulou, Ya Le +6
Most existing recommender systems focus primarily on matching users to content which maximizes user satisfaction on the platform. It is increasingly obvious, however, that content…
RecSim NG: Toward Principled Uncertainty Modeling for Recommender Ecosystems
Martin Mladenov, Chih-Wei Hsu, Vihan Jain +7
The development of recommender systems that optimize multi-turn interaction with users, and model the interactions of different agents (e.g., users, content providers, vendors) in…
Meta-Thompson Sampling
Branislav Kveton, Mikhail Konobeev, Manzil Zaheer +4
Efficient exploration in bandits is a fundamental online learning problem. We propose a variant of Thompson sampling that learns to explore better as it interacts with bandit insta…
Optimizing Long-term Social Welfare in Recommender Systems: A Constrained Matching Approach
Martin Mladenov, Elliot Creager, Omer Ben-Porat +3
Most recommender systems (RS) research assumes that a user's utility can be maximized independently of the utility of the other agents (e.g., other users, content providers). In re…
Meta-Learning Bandit Policies by Gradient Ascent
Branislav Kveton, Martin Mladenov, Chih-Wei Hsu +3
Most bandit policies are designed to either minimize regret in any problem instance, making very few assumptions about the underlying environment, or in a Bayesian sense, assuming…
Differentiable Bandit Exploration
Craig Boutilier, Chih-Wei Hsu, Branislav Kveton +3
Exploration policies in Bayesian bandits maximize the average reward over problem instances drawn from some distribution . In this work, we learn such policies for an…
RecSim: A Configurable Simulation Platform for Recommender Systems
Eugene Ie, Chih-wei Hsu, Martin Mladenov +5
We propose RecSim, a configurable platform for authoring simulation environments for recommender systems (RSs) that naturally supports sequential interaction with users. RecSim all…
Advantage Amplification in Slowly Evolving Latent-State Environments
Martin Mladenov, Ofer Meshi, Jayden Ooi +2
Latent-state environments with long horizons, such as those faced by recommender systems, pose significant challenges for reinforcement learning (RL). In this work, we identify and…
Empirical Bayes Regret Minimization
Chih-Wei Hsu, Branislav Kveton, Ofer Meshi +2
Most bandit algorithm designs are purely theoretical. Therefore, they have strong regret guarantees, but also are often too conservative in practice. In this work, we pioneer the i…