On pattern classification with Sammon's nonlinear mapping an experimental study

• Published in: Pattern recognition Vol. 31; no. 4; pp. 371 - 381
• Main Authors: Lerner, Boaz, Guterman, Hugo, Aladjem, Mayer, Dinsteint, Its'hak, Romem, Yitzhak
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01-04-1998; Elsevier Science
• Subjects: Algorithmics. Computability. Computer arithmetics; Applied sciences; Artificial intelligence; Chromosomes; Classification; Computer science; control theory; systems; Connectionism. Neural networks; Exact sciences and technology; Feature extraction; Multilayer perceptron; Neural networks; Pattern recognition. Digital image processing. Computational geometry; Sammon's mapping; Theoretical computing; Feature extraction; Multilayer perceptron; Neural networks; Sammon's mapping; Chromosomes; Classification; System architecture; Computer simulation; Network architecture; Chromosome; Mapping; Pattern recognition; Neural network; Experimental study; Algorithm; Perceptron; System performance; Data structure; Pattern extraction; Principal component analysis
• ISSN: 0031-3203; 1873-5142
• DOI: 10.1016/S0031-3203(97)00064-2

Sammon's mapping is conventionally used for exploratory data projection, and as such is usually inapplicable for classification. In this paper we apply a neural network (NN) implementation of Sammon's mapping to classification by extracting an arbitrary number of projections. The projection map and classification accuracy of the mapping are compared with those of the auto-associative NN (AANN), multilayer perceptron (MLP) and principal component (PC) feature extractor for chromosome data. We demonstrate that chromosome classification based on Sammon's (unsupervised) mapping is superior to the classification based on the AANN and PC feature extractor and highly comparable with that based on the (supervised) MLP. c 1998 Pattern Recognition Society.