Model-based contextual policy search for data-efficient generalization of robot skills

DOI10.1016/J.ARTINT.2014.11.005MaRDI QIDQ2407451zbMATH OpenOpenAlexWikidataFDO

Authors Andras Kupcsik, Marc Peter Deisenroth, Jan Peters, Loh Ai Poh, Prahlad Vadakkepat, Gerhard Neumann

Publication date 29 September 2017

Published in Artificial Intelligence (Search for Journal in Brave)

Full work available at URL http://eprints.lincoln.ac.uk/id/eprint/25774/1/Kupcsik_AIJ_2015.pdf

zbMATH Keywords

Gaussian processesreinforcement learningroboticsmovement primitivescontextual policy searchmodel-based policy searchrobot hockeyrobot skill generalizationrobot table tennis

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Artificial intelligence for robotics (68T40)

Recommendations

• Active contextual policy search
• Policy search for motor primitives in robotics
• Robot control policy transfer based on progressive neural network
• Hierarchical relative entropy policy search
• Towards motor skill learning for robotics

Cites work

• scientific article; zbMATH DE number 2107836 (Why is no real title available?)
• Simple statistical gradient-following algorithms for connectionist reinforcement learning

Cited in

(12)

• Editorial: Towards a science of integrated AI and robotics
• Hybrid control for learning motor skills
• Non-parametric policy search with limited information loss
• Model-based policy gradients with parameter-based exploration by least-squares conditional density estimation
• A survey of preference-based reinforcement learning methods
• scientific article; zbMATH DE number 6982305 (Why is no real title available?)
• Active contextual policy search
• Variational policy search using sparse Gaussian process priors for learning multimodal optimal actions
• Entropy regularization methods for parameter space exploration
• Using simulation to improve sample-efficiency of Bayesian optimization for bipedal robots
• Reinforcement learning in sparse-reward environments with hindsight policy gradients
• Robot control policy transfer based on progressive neural network

Describes a project that uses

Uses Software

• PILCO

This page was built for publication: Model-based contextual policy search for data-efficient generalization of robot skills

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2407451)