An Empirical Study On The Practical Impact Of Prior Beliefs Over Policy Types

Abstract

Many multiagent applications require an agent to learn quickly how to interact with previously unknown other agents. To address this problem, researchers have studied learning algorithms which compute posterior beliefs over a hypothesised set of policies, based on the observed actions of the other agents. The posterior belief is complemented by the prior belief, which specifies the subjective likelihood of policies before any actions are observed. In this paper, we present the first comprehensive empirical study on the practical impact of prior beliefs over policies in repeated interactions. We show that prior beliefs can have a significant impact on the long-term performance of such methods, and that the magnitude of the impact depends on the depth of the planning horizon. Moreover, our results demonstrate that automatic methods can be used to compute prior beliefs with consistent performance effects. This indicates that prior beliefs could be eliminated as a manual parameter and inst

Stats

• citations4
• S2 citations—
• github stars0
• HF likes0
• heat score5.24
• arxiv keyalbrecht2019an

Related papers

• On Convergence And Optimality Of Best-response Learning With Policy Types In Multiagent Systems (2019)0.00
• Reasoning About Unforeseen Possibilities During Policy Learning (2018)0.00
• Generalized Beliefs For Cooperative AI (2022)0.00
• Deep Interactive Bayesian Reinforcement Learning Via Meta-learning (2021)5.24
• Policy Evaluation In Decentralized Pomdps With Belief Sharing (2023)0.00
• Agent-pro: Learning To Evolve Via Policy-level Reflection And Optimization (2024)9.59
• Mechanisms For A No-regret Agent: Beyond The Common Prior (2020)0.00
• Practical Performative Policy Learning With Strategic Agents (2024)0.00