On the Modeling of Error Functions as High Dimensional Landscapes for Weight Initialization in Learning Networks

Next generation deep neural networks for classification hosted on embedded platforms will rely on fast, efficient, and accurate learning algorithms. Initialization of weights in learning networks has a great impact on the classification accuracy. In this paper we focus on deriving good initial weigh...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2016-07
Main Authors: Julius, Gopinath Mahale, Sumana, T, Adityakrishna, C S
Format: Article
Language:English
Subjects:
Algorithms
Classification
Error functions
Landscape
Machine learning
Mathematical models
Matrix theory
Modelling
Neural networks
Weight
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Next generation deep neural networks for classification hosted on embedded platforms will rely on fast, efficient, and accurate learning algorithms. Initialization of weights in learning networks has a great impact on the classification accuracy. In this paper we focus on deriving good initial weights by modeling the error function of a deep neural network as a high-dimensional landscape. We observe that due to the inherent complexity in its algebraic structure, such an error function may conform to general results of the statistics of large systems. To this end we apply some results from Random Matrix Theory to analyse these functions. We model the error function in terms of a Hamiltonian in N-dimensions and derive some theoretical results about its general behavior. These results are further used to make better initial guesses of weights for the learning algorithm.
ISSN:2331-8422