Doctor of Crosswise: Reducing Over-parametrization in Neural Networks

2019/05/24 by J. D. Curtó, I. C. Zarza, Curtó, J. D. +7
Computer Science · Mathematics · #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1905.10324

arxiv created 2020/04/17 · arxiv updated 2020/04/20

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.

Related

• Neural networks: from the perceptron to deep nets
• Fit without fear: remarkable mathematical phenomena of deep learning through the prism of interpolation
• Deep learning: Technical introduction
• The Modern Mathematics of Deep Learning
• Nearly Minimal Over-Parametrization of Shallow Neural Networks