The adaptive-subspace self-organizing map (AS-SOM) proposed by Kohonen is a
recent development in self-organizing map (SOM) computation. In this paper
, we propose a method to realize ASSOM using a neural learning algorithm in
nonlinear autoencoder networks. Our method has the advantage of numerical
stability. We have applied our ASSOM model to build a modular classificatio
n system for handwritten digit recognition. Ten ASSOM modules are used to c
apture different features in the ten classes of digits. When a test digit i
s presented to all the modules, each module provides a reconstructed patter
n and the system outputs a class label by comparing the ten reconstruction
errors, Our experiments show promising results. For relatively small size m
odules, the classification accuracy reaches 99.3% on the training set and o
ver 97% on the testing set.