A probabilistic model for evaluation of neural network classifiers

• Published in: Pattern recognition Vol. 25; no. 10; pp. 1241 - 1251
• Main Authors: Musavi, M.T., Chan, K.H., Hummels, D.M., Kalantri, K., Ahmed, W.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.10.1992; Elsevier Science
• Subjects: Applied sciences; Artificial intelligence; Computer science; control theory; systems; Exact sciences and technology; Generalization; Gram-Schmidt orthogonalization; Neural networks; Pattern recognition. Digital image processing. Computational geometry; Probabilistic evaluation model; Gram-Schmidt orthogonalization; Generalization; Neural networks; Probabilistic evaluation model; Learning; Performance evaluation; Monte Carlo method; Error function; Probabilistic model; Classifier; Pattern recognition; Neural network; Computer system; Artificial intelligence; Implementation
• ISSN: 0031-3203; 1873-5142
• DOI: 10.1016/0031-3203(92)90025-E

A technique for evaluation of the generalization ability in artificial neural network (ANN) classifiers is presented. A probabilistic input model is proposed to account for all possible input ranges. The expected value of a square error function over the defined input range is taken as a measure of generalization ability. The minimization of the error function outlines the boundary of the decision region for a minimum error neural network (MENN) classifier. Two essential elements for carrying out the proposed technique are the estimation of the input density and numerical integration. A non-parametric method is used to locally estimate the distribution around each training pattern. The Monte Carlo method has been used for numerical integration. The evaluation technique was tested for measuring the generalization ability of back propagation (BP), radial basis function (RBF), probabilistic neural network (PNN) and MENN classifiers for different problems.