Closed-loop deep learning: generating forward models with
back-propagation
S. Daryanavard, and B. Porr. (2020)cite arxiv:2001.02970Comment: 13 pages, 6 figures.
Abstract
A reflex is a simple closed loop control approach which tries to minimise an
error but fails to do so because it will always react too late. An adaptive
algorithm can use this error to learn a forward model with the help of
predictive cues. For example a driver learns to improve their steering by
looking ahead to avoid steering in the last minute. In order to process complex
cues such as the road ahead deep learning is a natural choice. However, this is
usually only achieved indirectly by employing deep reinforcement learning
having a discrete state space. Here, we show how this can be directly achieved
by embedding deep learning into a closed loop system and preserving its
continuous processing. We show specifically how error back-propagation can be
achieved in z-space and in general how gradient based approaches can be
analysed in such closed loop scenarios. The performance of this learning
paradigm is demonstrated using a line-follower both in simulation and on a real
robot that show very fast and continuous learning.
Description
[2001.02970] Closed-loop deep learning: generating forward models with back-propagation
%0 Generic
%1 daryanavard2020closedloop
%A Daryanavard, Sama
%A Porr, Bernd
%D 2020
%K ClosedLoop DeepLearning
%T Closed-loop deep learning: generating forward models with
back-propagation
%U http://arxiv.org/abs/2001.02970
%X A reflex is a simple closed loop control approach which tries to minimise an
error but fails to do so because it will always react too late. An adaptive
algorithm can use this error to learn a forward model with the help of
predictive cues. For example a driver learns to improve their steering by
looking ahead to avoid steering in the last minute. In order to process complex
cues such as the road ahead deep learning is a natural choice. However, this is
usually only achieved indirectly by employing deep reinforcement learning
having a discrete state space. Here, we show how this can be directly achieved
by embedding deep learning into a closed loop system and preserving its
continuous processing. We show specifically how error back-propagation can be
achieved in z-space and in general how gradient based approaches can be
analysed in such closed loop scenarios. The performance of this learning
paradigm is demonstrated using a line-follower both in simulation and on a real
robot that show very fast and continuous learning.
@misc{daryanavard2020closedloop,
abstract = {A reflex is a simple closed loop control approach which tries to minimise an
error but fails to do so because it will always react too late. An adaptive
algorithm can use this error to learn a forward model with the help of
predictive cues. For example a driver learns to improve their steering by
looking ahead to avoid steering in the last minute. In order to process complex
cues such as the road ahead deep learning is a natural choice. However, this is
usually only achieved indirectly by employing deep reinforcement learning
having a discrete state space. Here, we show how this can be directly achieved
by embedding deep learning into a closed loop system and preserving its
continuous processing. We show specifically how error back-propagation can be
achieved in z-space and in general how gradient based approaches can be
analysed in such closed loop scenarios. The performance of this learning
paradigm is demonstrated using a line-follower both in simulation and on a real
robot that show very fast and continuous learning.},
added-at = {2021-11-29T11:35:47.000+0100},
author = {Daryanavard, Sama and Porr, Bernd},
biburl = {https://www.bibsonomy.org/bibtex/2f28eec4349a6bb53f3413c8ba6c8c6e0/annakrause},
description = {[2001.02970] Closed-loop deep learning: generating forward models with back-propagation},
interhash = {4f258013905723e14f24da29ca6e8fb1},
intrahash = {f28eec4349a6bb53f3413c8ba6c8c6e0},
keywords = {ClosedLoop DeepLearning},
note = {cite arxiv:2001.02970Comment: 13 pages, 6 figures},
timestamp = {2021-11-29T11:35:47.000+0100},
title = {Closed-loop deep learning: generating forward models with
back-propagation},
url = {http://arxiv.org/abs/2001.02970},
year = 2020
}