Online learning of neural networks based on a model-free control algorithm
Systems and Control
2021-08-31 v3 Systems and Control
Optimization and Control
Abstract
We explore the possibilities of using a model-free-based control law in order to train artificial neural networks. In the supervised learning context, we consider the problem of tuning the synaptic weights as a feedback control tracking problem where the control algorithm adjusts the weights online according to the input-output training data set of the neural network. Numerical results illustrate the dynamical learning process and an example of classifier that show very promising properties of our proposed approach.
Cite
@article{arxiv.1905.02230,
title = {Online learning of neural networks based on a model-free control algorithm},
author = {Loïc Michel},
journal= {arXiv preprint arXiv:1905.02230},
year = {2021}
}
Comments
8 pages, 4 figures, Accepted to the APMS 2021 conference