Add to ORKGin AI and operations research focuses on solving a single problem. However, in practice, AI agents often exist over a longer period of time, during which they may be required to solve several related tasks. This type of scenario has motivated a significant amount of recent research in knowledge transfer methods for MDPs. The idea is to allow an agent to continue to re-use the expertise accumulated while solving past tasks over its lifetime. Several approaches for knowledge transfer in MDPs have been proposed, and a compre-hensive survey is provided in [5]. An extended version of this paper with all the proofs, more experimental results and related work is provided in [1]. We focus on transferring knowledge in MDPs that are specified fully b...
Artificial Intelligence (AI) is the branch of the Computer Science field that tries to imbue intelli...
Thesis (Ph.D.), Computer Science, Washington State UniversityReinforcement learning (RL) has had man...
We consider the problem of transferring a learned optimal policy between MDPs. We describe a method ...
Knowledge transfer has been suggested as a useful approach for solving large Markov Decision Process...
Decision makers that employ state abstrac-tion (or state aggregation) usually find solu-tions faster...
We consider how to transfer knowledge from previous tasks (MDPs) to a current task in long-lived and...
We present algorithms to effectively represent a set of Markov decision processes (MDPs), whose opti...
As any other problem solving task that employs search, AI Planning needs heuristics to efficiently g...
In this article, we discuss the problem of transferring search heuristics from one planner to anothe...
This thesis concerns sample-efficient embodied machine learning. Machine learning success in sequent...
The goal of transfer is to use knowledge obtained by solving one task to improve a robot’s (or softw...
Transfer learning can improve the reinforcement learn-ing of a new task by allowing the agent to reu...
The MDP formalism and its variants are usually used to control the state of a system through an agen...
This paper describes an algorithm for model checking a fragment of the logic of knowledge and probab...
Reinforcement learning algorithms are very effective at learning policies (mappings from states to a...
Artificial Intelligence (AI) is the branch of the Computer Science field that tries to imbue intelli...
Thesis (Ph.D.), Computer Science, Washington State UniversityReinforcement learning (RL) has had man...
We consider the problem of transferring a learned optimal policy between MDPs. We describe a method ...
Knowledge transfer has been suggested as a useful approach for solving large Markov Decision Process...
Decision makers that employ state abstrac-tion (or state aggregation) usually find solu-tions faster...
We consider how to transfer knowledge from previous tasks (MDPs) to a current task in long-lived and...
We present algorithms to effectively represent a set of Markov decision processes (MDPs), whose opti...
As any other problem solving task that employs search, AI Planning needs heuristics to efficiently g...
In this article, we discuss the problem of transferring search heuristics from one planner to anothe...
This thesis concerns sample-efficient embodied machine learning. Machine learning success in sequent...
The goal of transfer is to use knowledge obtained by solving one task to improve a robot’s (or softw...
Transfer learning can improve the reinforcement learn-ing of a new task by allowing the agent to reu...
The MDP formalism and its variants are usually used to control the state of a system through an agen...
This paper describes an algorithm for model checking a fragment of the logic of knowledge and probab...
Reinforcement learning algorithms are very effective at learning policies (mappings from states to a...
Artificial Intelligence (AI) is the branch of the Computer Science field that tries to imbue intelli...
Thesis (Ph.D.), Computer Science, Washington State UniversityReinforcement learning (RL) has had man...
We consider the problem of transferring a learned optimal policy between MDPs. We describe a method ...