Share Email Print
cover

Proceedings Paper

Geometry of decision boundaries of neural networks
Author(s): Chulhee Lee; Ohjae Kwon; Eunsuk Jung
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In this paper, we provide a thorough analysis of decision boundaries of neural networks when they are used as a classifier. It has been shown that the classifying mechanism of the neural network can be divided into two parts: dimension expansion by hidden neurons and linear decision boundary formation by output neurons. In this paradigm, the input data is first warped into a higher dimensional space by the hidden neurons and the output neurons draw linear decision boundaries in the expanded space (hidden neuron space). We also note that the decision boundaries in the hidden neuron space are not completely independent. This dependency of decision boundaries is extended to multiclass problems, providing a valuable insight into formation of decision boundaries in the hidden neuron space. This analysis provides a new understanding of how neural networks construct complex decision boundaries and explains how different sets of weights may prove similar results.

Paper Details

Date Published: 13 November 2001
PDF: 13 pages
Proc. SPIE 4471, Algorithms and Systems for Optical Information Processing V, (13 November 2001); doi: 10.1117/12.449334
Show Author Affiliations
Chulhee Lee, Yonsei Univ. (South Korea)
Ohjae Kwon, Yonsei Univ. (South Korea)
Eunsuk Jung, Yonsei Univ. (South Korea)


Published in SPIE Proceedings Vol. 4471:
Algorithms and Systems for Optical Information Processing V
Bahram Javidi; Demetri Psaltis, Editor(s)

© SPIE. Terms of Use
Back to Top