A diagnosis system for multi class primary headaches using ant miner plus algorithm

Abstract

This article proposes a diagnosis system for dealing with primary headache classification using variant of ant colony optimization algorithm for classification. The diagnosis system proposed is an expert system completely designed with information gathered from patients suffering from headaches. Randomly chosen patients who are visiting various hospitals in India are addressed to collect data for expert system. They are provided with detailed questionnaire about the symptoms to analyze the headache types. With help of neurologist the headaches are compared with the results of ant colony optimization algorithms for classification. Ant miner plus algorithm is enforced for better results. The algorithm is observed for its accuracy levels of classification and is analyzed. This article addresses 4 types of headaches as multi-class classification problem with an in detailed report of symptoms gathered from patients. This kind of expert system is useful to neurologists to track symptoms of their patients and to provide mediation.