Doctor of Crosswise: Reducing Over-parametrization in Neural Networks
Machine Learning
2020-04-20 v3 Machine Learning
Abstract
Dr. of Crosswise proposes a new architecture to reduce over-parametrization in Neural Networks. It introduces an operand for rapid computation in the framework of Deep Learning that leverages learned weights. The formalism is described in detail providing both an accurate elucidation of the mechanics and the theoretical implications.
Cite
@article{arxiv.1905.10324,
title = {Doctor of Crosswise: Reducing Over-parametrization in Neural Networks},
author = {J. D. Curtó and I. C. Zarza and Kris Kitani and Irwin King and Michael R. Lyu},
journal= {arXiv preprint arXiv:1905.10324},
year = {2020}
}