Summary: Comparison of Kernel Estimators, Perceptrons, and Radial­Basis
Functions for OCR and Speech Classification
Ethem Alpaydin, Fikret G¨urgen
Department of Computer Engineering, Bo–gazi¸ci University, TR­80815 Istanbul Turkey
alpaydin@boun.edu.tr
Neural Computing & Applications (1995) 3: 38­49
c
fl1995 Springer­Verlag London Limited
Abstract---We compare kernel estimators, single and
multi­layered perceptrons and radial­basis functions for
the problems of classification of handwritten digits and
speech phonemes. By taking two different applications
and employing many techniques, we report here a two­
dimensional study whereby a domain­independent assess­
ment of these learning methods can be possible. We con­
sider a feed­forward network with one hidden layer. As
examples of the local methods, we use kernel estimators
like k­nearest neighbor (k­nn), Parzen windows, gener­
alized k­nn, and Grow and Learn (Condensed Nearest
Neighbor). We have also considered fuzzy k­nn due to

Source: Alpaydýn, Ethem - Department of Computer Engineering, Bogaziçi University

Collections: Computer Technologies and Information Sciences