Skip to Main content Skip to Navigation
Journal articles

Respective Advantages and Disadvantages of Model-based and Model-free Reinforcement Learning in a Robotics Neuro-inspired Cognitive Architecture

Erwan Renaudo 1, 2, * Benoît Girard 1, 2 Raja Chatila 1, 2 Mehdi Khamassi 1, 2
* Corresponding author
2 AMAC
ISIR - Institut des Systèmes Intelligents et de Robotique
Abstract : Combining model-based and model-free reinforcement learning systems in robotic cognitive architectures appears as a promising direction to endow artificial agents with flexibility and decisional autonomy close to mammals. In particular, it could enable robots to build an internal model of the environment, plan within it in response to detected environmental changes, and avoid the cost and time of planning when the stability of the environment is recognized as enabling habit learning. However, previously proposed criteria for the coordination of these two learning systems do not scale up to the large, partial and uncertain models autonomously learned by robots. Here we precisely analyze the performances of these two systems in an asynchronous robotic simulation of a cube-pushing task requiring a permanent trade-off between speed and accuracy. We propose solutions to make learning successful in these conditions. We finally discuss possible criteria for their efficient coordination within robotic cognitive architectures.
Document type :
Journal articles
Complete list of metadata

Cited literature [14 references]  Display  Hide  Download

https://hal.sorbonne-universite.fr/hal-01250157
Contributor : Gestionnaire Hal-Upmc <>
Submitted on : Monday, January 4, 2016 - 1:24:27 PM
Last modification on : Friday, March 19, 2021 - 3:32:54 AM
Long-term archiving on: : Friday, April 15, 2016 - 4:01:59 PM

File

1-s2.0-S1877050915036558-main....
Publication funded by an institution

Licence


Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0 International License

Identifiers

Citation

Erwan Renaudo, Benoît Girard, Raja Chatila, Mehdi Khamassi. Respective Advantages and Disadvantages of Model-based and Model-free Reinforcement Learning in a Robotics Neuro-inspired Cognitive Architecture. Procedia Computer Science, Elsevier, 2015, 71, pp.178-184. ⟨10.1016/j.procs.2015.12.194⟩. ⟨hal-01250157⟩

Share

Metrics

Record views

659

Files downloads

1120