Product Units: A Computationally Powerful and Biologically Plausible Extension to Backpropagation Networks
Abstract
We introduce a new form of computational unit for feedforward learning networks of the backpropagation type. Instead of calculating a weighted sum this unit calculates a weighted product, where each input is raised to a power determined by a variable weight. Such a unit can learn an arbitrary polynomial term, which would then feed into higher level standard summing units. We show how learning operates with product units, provide examples to show their efficiency for various types of problems, and argue that they naturally extend the family of theoretical feedforward net structures. There is a plausible neurobiological interpretation for one interesting configuration of product and summing units.
Cite
Text
Durbin and Rumelhart. "Product Units: A Computationally Powerful and Biologically Plausible Extension to Backpropagation Networks." Neural Computation, 1989. doi:10.1162/NECO.1989.1.1.133Markdown
[Durbin and Rumelhart. "Product Units: A Computationally Powerful and Biologically Plausible Extension to Backpropagation Networks." Neural Computation, 1989.](https://mlanthology.org/neco/1989/durbin1989neco-product/) doi:10.1162/NECO.1989.1.1.133BibTeX
@article{durbin1989neco-product,
title = {{Product Units: A Computationally Powerful and Biologically Plausible Extension to Backpropagation Networks}},
author = {Durbin, Richard and Rumelhart, David E.},
journal = {Neural Computation},
year = {1989},
pages = {133-142},
doi = {10.1162/NECO.1989.1.1.133},
volume = {1},
url = {https://mlanthology.org/neco/1989/durbin1989neco-product/}
}