Artificial Neural Network Topology Optimization Using Bees Algorithm For Classification Of Melanoma

This paper develops an automatically melanoma classification system, in order to presents an approach toArtificial Neural Network topology optimization by using Bees Algorithm. The best network architecture isfound for melanoma database classification. Part of principal this system is an ANN that is able to classifycases. The database is used in this paper that contains 290 images and each of them is defined by 12parameters, the constitutive first 11 values give a report for features extracted from images and lastparameter represents the nature of the skin lesion. In this database, two classes are considered: melanomaand non-melanoma. Results showed that this system can be achieved in a total accuracy of 91.92%.