Application of Max-Plus based Neural Networks to Dermatology Disease Classification.

Various computational intelligence-based techniques have been developed to help medical decisions using data based on the expert knowledge of doctors. In this paper, Max-Plus based neural networks are applied to the problem of dermatology disease classification. The advantages of Max-Plus based neural networks are mainly correspond to high speed/parallel processing, analyzing information in terms of ordered structure, and treating only discreteinformation with no quantization error. To optimize the connection weights of Max-Plus based neural networks, the memetic algorithm is considered rather than the gradient-based learning methods because of its poor convergence properties. To verify the applicability of the proposed method, dermatology data set available on the Machine Learning Repository site at the University of California at Irvine is used. [ABSTRACT FROM AUTHOR]