Generalized Non-Reducible Descriptors

This paper provides a generalization of non-reducible descriptors by extending the concept of distance between patterns of di erent classes. Generalized non-reducible descriptors are used in supervised pattern recognition problems where the feature vectors consist of Boolean variables. Generalized non-reducible descriptors are expressed as conjunctions and correspond to syndromes in medical diagnosis. Generalized non-reducible descriptors minimize the number of operations in the decision rules. A mathematical model to construct generalized non-reducible descriptors, a computational procedure, and numerical examples are discussed. ? 2004 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.

• Type: Conference
• Year: 2000
• Where: ICPR
• Authors: Ventzeslav Valev, Bülent Sankur, Petia Radeva
• Keywords: ICPR 2000, Non-reducible Descriptors, Pattern Recognition, Supervised Pattern Recognition