Fast Learning with Predictive Forward Models
Abstract
A method for transforming performance evaluation signals distal both in space and time into proximal signals usable by supervised learning algo(cid:173) rithms, presented in [Jordan & Jacobs 90], is examined. A simple obser(cid:173) vation concerning differentiation through models trained with redundant inputs (as one of their networks is) explains a weakness in the original architecture and suggests a modification: an internal world model that encodes action-space exploration and, crucially, cancels input redundancy to the forward model is added. Learning time on an example task, cart(cid:173) pole balancing, is thereby reduced about 50 to 100 times.
Cite
Text
Brody. "Fast Learning with Predictive Forward Models." Neural Information Processing Systems, 1991.Markdown
[Brody. "Fast Learning with Predictive Forward Models." Neural Information Processing Systems, 1991.](https://mlanthology.org/neurips/1991/brody1991neurips-fast/)BibTeX
@inproceedings{brody1991neurips-fast,
title = {{Fast Learning with Predictive Forward Models}},
author = {Brody, Carlos},
booktitle = {Neural Information Processing Systems},
year = {1991},
pages = {563-570},
url = {https://mlanthology.org/neurips/1991/brody1991neurips-fast/}
}