Integrating Reinforcement Learning and Declarative Programming to Learn Causal Laws in Dynamic Domains

Sridharan, Mohan; Rainge, Sarah

doi:10.1007/978-3-319-11973-1_33

Cited by 4 publications

(2 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This paper builds on prior work in (a) combining declarative programming and probabilistic graphical models for planning and diagnostics in robotics [1], [2], [3]; (b) combining declarative programming with RL for heuristic discovery of axioms in simplistic domains [28]; and (c) introducing relational representations for more efficient discovery of axioms [4], [29]. Here, we use ASP-based reasoning, a modified relational representation, and a sampling-based algorithm for efficiently identifying candidate axioms and generalizing across axiom instances.…”

Section: Related Workmentioning

confidence: 99%

Should I do that? using relational reinforcement learning and declarative programming to discover domain axioms

Sridharan

Meadows

2016

2016 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)

Self Cite

View full text Add to dashboard Cite

Robots assisting humans in complex domains need the ability to represent, reason with, and learn from, different descriptions of incomplete domain knowledge and uncertainty. This paper focuses on the challenge of incrementally and interactively discovering previously unknown axioms governing domain dynamics, and describes an architecture that integrates declarative programming and relational reinforcement learning to address this challenge. Answer Set Prolog (ASP), a declarative programming paradigm, is used to represent and reason with incomplete domain knowledge for planning and diagnostics. For any given goal, unexplained failure of plans created by ASP-based inference is taken to indicate the existence of unknown domain axioms. The task of discovering these axioms is formulated as a reinforcement learning problem, and a relational representation is used to incrementally generalize from specific axioms identified over time. These generic axioms are then added to the ASP-based representation for subsequent inference. The architecture's capabilities are demonstrated and evaluated in two domains, Blocks World and Robot Butler.

show abstract

Section: Related Workmentioning

confidence: 99%

Should I do that? using relational reinforcement learning and declarative programming to discover domain axioms

Sridharan

Meadows

2016

2016 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)

Self Cite

View full text Add to dashboard Cite

show abstract

“…When world model is unavailable, SDM can be realized using RL. People have developed algorithms for integrating KRR and RL methods (Leonetti, Iocchi, and Stone 2016;Yang et al 2018;Lyu and others 2019;Sridharan and Rainge 2014;Griffith, Subramanian, and others 2013;Illanes et al 2019). Among them, Leonetti, Iocchi, and Stone used declarative action knowledge to help an agent to select only the reasonable actions in RL exploration.…”

Section: Introductionmentioning

confidence: 99%

Learning and Reasoning for Robot Sequential Decision Making under Uncertainty

Amiri¹,

Shirazi²,

Zhang³

2020

AAAI

View full text Add to dashboard Cite

Robots frequently face complex tasks that require more than one action, where sequential decision-making (sdm) capabilities become necessary. The key contribution of this work is a robot sdm framework, called lcorpp, that supports the simultaneous capabilities of supervised learning for passive state estimation, automated reasoning with declarative human knowledge, and planning under uncertainty toward achieving long-term goals. In particular, we use a hybrid reasoning paradigm to refine the state estimator, and provide informative priors for the probabilistic planner. In experiments, a mobile robot is tasked with estimating human intentions using their motion trajectories, declarative contextual knowledge, and human-robot interaction (dialog-based and motion-based). Results suggest that, in efficiency and accuracy, our framework performs better than its no-learning and no-reasoning counterparts in office environment.

show abstract

Can I Do That? Discovering Domain Axioms Using Declarative Programming and Relational Reinforcement Learning

Sridharan

Devarakonda

Gupta

2016

Autonomous Agents and Multiagent Systems

View full text Add to dashboard Cite

Integrating Reinforcement Learning and Declarative Programming to Learn Causal Laws in Dynamic Domains

Cited by 4 publications

References 11 publications

Should I do that? using relational reinforcement learning and declarative programming to discover domain axioms

Should I do that? using relational reinforcement learning and declarative programming to discover domain axioms

Learning and Reasoning for Robot Sequential Decision Making under Uncertainty

Can I Do That? Discovering Domain Axioms Using Declarative Programming and Relational Reinforcement Learning

Contact Info

Product

Resources

About