Please use this identifier to cite or link to this item:
Title: Training binary perceptron by clipping
Authors: SCHIETSE, Jan
BOUTEN, Marcus 
VAN DEN BROECK, Christian 
Issue Date: 1995
Source: Europhysics letters, 32(3). p. 279-284
Abstract: A teacher perceptron T with N binary components provides the classification of a set of p randomly chosen training examples. Several algorithms are available that use this information to select a student perceptron J with continuous components Ji. The purpose is to maximize the overlap R = T·J/(|T| |J|), or to minimize the corresponding generalization error ε = (1/π) arccos R. In view of the binary nature of the components of the teacher, one might expect that a lower error can be achieved by working with the clipped version of the student vector, namely the vector with components sign (Ji). It turns out that this is not always the case. In this letter we calculate the overlap for a vector with components f(Ji), where f can be any odd function of its argument, as a function of the overlap R. We show that the optimal choice of f is a hyperbolic tangent f(x) = th ((R/(1 - R2)x)). The corresponding generalization error can go to zero exponentially fast in a2, for a large (a = p/N).
Document URI:
DOI: 10.1209/0295-5075/32/3/015
Type: Journal Contribution
Appears in Collections:Research publications

Show full item record


checked on Sep 2, 2020


checked on May 22, 2022

Page view(s)

checked on May 23, 2022

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.