Evolving Fuzzy C-Means: An intelligent technique for efficient diagnosis of children mental retardation level from databases with missing values

In psychopathological diagnosis, a correct classification of mental retardation level is needed to choose the better treatment for rehabilitation and to assure a quality of life suitable for the specific patient condition. In order to meet this need we studied a technique that allows to perform automatic diagnoses efficiently and reliably and at the same time is easy for psychotherapists to use. In this paper we present a new approach which integrates two well-known computational intelligence techniques: Fuzzy CMeans and Genetic Algorithms. This integration allows automatic generation of an optimized classifier for efficient recognition of mental retardation level starting from databases with missing values. The results obtained with this simple technique are encouraging: an empirical test on our database of patients showed the efficiency of the simple integrated algorithm, which also gives a great deal of useful information for diagnostic purposes.