The QR decomposition for radial neural networks

07/06/2021

∙

by   Iordan Ganev, et al.

∙

0

∙

share

We provide a theoretical framework for neural networks in terms of the representation theory of quivers, thus revealing symmetries of the parameter space of neural networks. An exploitation of these symmetries leads to a model compression algorithm for radial neural networks based on an analogue of the QR decomposition. A projected version of backpropogation on the original model matches usual backpropogation on the compressed model.