Black-Box Data-efficient Policy Search for Robotics

Chatzilygeroudis, Konstantinos; Rama, Roberto; Kaushik, Rituraj; Goepp, Dorian; Vassiliades, Vassilis; Mouret, Jean-Baptiste

doi:10.48550/arxiv.1703.07261

Search citation statements

Order By: Relevance

Paper Sections

Select...

Perspective and Structure Of The Survey1

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2019

Publication Types

Select...

Article1

Relationship

Self Cite0

Independent1

Authors

Journals

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Sections 2 to 5, we present Solutions 1 to 4 above in more detail, showing how the corresponding methods are implemented in various policy search algorithms. We do not cover Solution 5 and refer readers to (Chatzilygeroudis et al, 2017) for a recent presentation of these model-based policy O is an outcome space (see Section 4) and X × U is the state and action space (see Section 5). Algorithms not covered in (Deisenroth et al, 2013) have a lighter (green) background.…”

Section: Perspective and Structure Of The Surveymentioning

confidence: 99%

Policy search in continuous action domains: An overview

Sigaud

Stulp

2019

Neural Networks

View full text Add to dashboard Cite

Continuous action policy search is currently the focus of intensive research, driven both by the recent success of deep reinforcement learning algorithms and the emergence of competitors based on evolutionary algorithms. In this paper, we present a broad survey of policy search methods, providing a unified perspective on very different approaches, including also Bayesian Optimization and directed exploration methods. The main message of this overview is in the relationship between the families of methods, but we also outline some factors underlying sample efficiency properties of the various approaches.In the context of robotics, sample efficiency is a key concern. There are three aspects to sample ef-arXiv:1803.04706v5 [cs.LG]

show abstract

Section: Perspective and Structure Of The Surveymentioning

confidence: 99%