Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. IO, NO. 2 . MARCH 1988 277 The Capacity of Multilevel Threshold Functions
 

Summary: IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. IO, NO. 2 . MARCH 1988 277
The Capacity of Multilevel Threshold Functions
SVERRIR 'OLAFSSON AND YASER S. ABU-MOSTAFA
Abstract-The investigation of information capacity and computa-
tion power of threshold elements is motivated by the recent trend of
neural-network approaches to pattern recognition. In this correspon-
dence, we estimate lower and upper bounds for the capacity of multi-
level threshold elements, using two essentially different enumeration
techniques. The results correct a previously published estimate and
indicate that adding threshold levels enhances the capacity more than
adding variables.
Index Terms-Classifier, information capacity, multilevelthreshold
element, neural networks, parallel hyperplanes, pattern recognition.
I. INTRODUCTION
Threshold functions have somewhat regained interest in recent
years. After considerable activity in the early 1960's, the subject
all but vanished in the following decade. Current interest is spumed,
among other things, by the unexpected difficulty of realizing effi-
cient pattern recognition schemes on conventional digital com-
puters. As an alternative, threshold functions offer a reasonable

  

Source: Abu-Mostafa, Yaser S. - Department of Mechanical Engineering & Computer Science Department, California Institute of Technology

 

Collections: Computer Technologies and Information Sciences