A PARALLEL NETWORK OF MODIFIED 1-NN AND K-NN CLASSIFIERS - APPLICATION TO REMOTE-SENSING IMAGE CLASSIFICATION

A parallel network of modified 1-NN classifiers and R-NN classifiers i s described and compared with a standard k-NN classifier. All the comp onent classifiers decide between two classes only. The number of all p ossible pairs of classes determines the number of the component classi fiers. The global decision is formed by voting of all the component cl assifiers. Each of the component classifiers operates as follows. For each class i a certain area A(i) is constructed in such a way that are a A(i) covers all training samples from the class i and possibly a sma ll number of training samples from other classes. In the classificatio n phase, if a sample lies outside of all areas A(i), then the classifi cation is refused. When it belongs only to one of the areas Ai, then t he classification is performed by the 1-NN rule. Samples that lie in a n overlapping area of some A(i) are classified by the k-NN rule. Such a classification rule, in this paper called a combined (I-NN, k-NN) ru le, is used by all component classifiers. Two feature selection sessio ns are recommended for each of the component classifiers: one to minim ize the size of the overlapping areas and another to minimize the erro r rate for the k-NN rule. The aim of this work is to create a classifi er with improved performance compared to the standard k-NN rule. It is shown that the replacement of the k-NN rule by the combined (1-NN, k- NN) rule reduces computing time required for classification while the parallelization of the classifier structure decreases the error rate. The effectiveness of the proposed approach was verified on a real data set of 5 classes, 15 features and 8839 samples which was derived from a couple of multisensorial remote-sensing images. (C) 1998 Elsevier S cience B.V.