Bridging the gap between reinforcement learning and knowledge representation: a logical off- and on-policy framework

DOI10.1007/978-3-642-22152-1_40MaRDI QIDQ3011967zbMATH OpenOpenAlexFDO

Authors Emad W. Saad

Publication date 29 June 2011

Published in Lecture Notes in Computer Science (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1012.1552

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Analysis of algorithms and problem complexity (68Q25) Computational difficulty of problems (lower bounds, completeness, difficulty of approximation, etc.) (68Q17) Knowledge representation (68T30) Logic in artificial intelligence (68T27)

Abstract: Knowledge Representation is important issue in reinforcement learning. In this paper, we bridge the gap between reinforcement learning and knowledge representation, by providing a rich knowledge representation framework, based on normal logic programs with answer set semantics, that is capable of solving model-free reinforcement learning problems for more complex do-mains and exploits the domain-specific knowledge. We prove the correctness of our approach. We show that the complexity of finding an offline and online policy for a model-free reinforcement learning problem in our approach is NP-complete. Moreover, we show that any model-free reinforcement learning problem in MDP environment can be encoded as a SAT problem. The importance of that is model-free reinforcement

Recommendations

• The effect of representation and knowledge on goal-directed exploration with reinforcement-learning algorithms
• Relational reinforcement learning
• Inductive Logic Programming
• Machine Learning: ECML 2003
• scientific article; zbMATH DE number 5558275

Cites work

• scientific article; zbMATH DE number 1216123 (Why is no real title available?)
• scientific article; zbMATH DE number 1315585 (Why is no real title available?)
• A new approach to hybrid probabilistic logic programs
• ASSAT: computing answer sets of a logic program by SAT solvers
• Answer set programming based on propositional satisfiability
• Contingent planning under uncertainty via stochastic satisfiability
• Domain-dependent knowledge in answer set planning
• Elevator group control using multiple reinforcement learning agents
• Inductive Logic Programming
• On the Relationship between Hybrid Probabilistic Logic Programs and Stochastic Satisfiability
• Practical solution techniques for first-order MDPs
• Probabilistic Reasoning by SAT Solvers
• Reasoning about actions with sensing under qualitative and probabilistic uncertainty
• SAT-based planning in complex domains: Concurrency, constraints and nondeterminism

Cited in

(4)

• scientific article; zbMATH DE number 5584479 (Why is no real title available?)
• Machine Learning: ECML 2003
• Low-Rank Representation of Reinforcement Learning Policies
• The effect of representation and knowledge on goal-directed exploration with reinforcement-learning algorithms

Describes a project that uses

Uses Software

• ASSAT

This page was built for publication: Bridging the gap between reinforcement learning and knowledge representation: a logical off- and on-policy framework

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3011967)