Gdswlyh 5 Hmhfw 2 Swlrq Iru / 94 & Odvvlilhuv

$EVWUDFW A reject rule devised for a neural classifier based on the Learning Vector Quantization (LVQ) paradigm is presented. The reject option is carried out adaptively to the specific application domain. It is assumed that a performance function P is defined which, taking into account the requirements of a given application expressed in terms of classification, misclassification and reject costs, evaluates the quality of the classification. Under this assumption the optimal reject threshold value, determining the best trade-off between reject rate and misclassification rate, is the one for which the function P reaches its absolute maximum. Implementation and performance of the rule are illustrated.