Permanent disability classification using hybrid neuro-logistic regression models

The social security administrations have to evaluate the degree of disability to offer compensation to those workers who suffer from a continuous alteration of health preventing them from continuing their work. Thanks to the accurate model of classification of disability presented in this paper, it is possible to obtain an approximation of the expected result for each case of disability prior to an individualized evaluation. In this paper, we introduce a novel model for classification of permanent disability, based on the hybridization of a standard logistic regression with Product Unit (PU) neural networks and Radial Basis Function (RBF) networks. The proposed techniques are shown to perform better than other existing Statistical and Artificial Intelligence methods.