An Intelligent Diagnostic System for Detection of Hepatitis using Multi-Layer Perceptron and Colonial Competitive Algorithm

This article proposes an intelligent diagnostic system for the diagnosis of Hepatitis based on a new algorithm including MLP and ICA which is more accurate and faster than the similar algorithms in terms of performance. At first, colonial competitive algorithm seeks to find the best solution in neural network training, then the MLP will be designed which can intelligently diagnose Hepatitis. Providing a certain solution and the ability to analyze complex, large-scale problems are among the advantages of this algorithm over similar optimization algorithm in diagnosis of Hepatitis. For neural network training and sample data testing 100 and 55 sample data were used respectively. Taken from UCI database, the data were applied to the system, revealing the effectiveness of the proposed algorithm in diagnosis of Hepatitis with less than 5% error.