HEART ATTACK DETECTION BY IMPROVING DATA MINING USING ACO TECHNIQUE

Abstract

Heart disease is a term that assigns a large number of medical conditions to the heart. These medical conditions describe abnormal health conditions which directly affect the heart and all its parts. There are many types of heart disease that can be considered congenital heart disease here, cardiac failure coronary heart disease. Based on the identified risk, we identify the maximum value of pheromones; The maximum value of pheromone is a combination of weight and risk level. The next step of ant is to find the maximum value of the pheromone because the ratio of the speed of the sensitive ant and its properties will change. With this approach, the number of pieces can be managed through the ACO parameter. In this research, we provide an efficient approach that is based on data mining and ant colony optimization techniques (ACO) for predicting cardiovascular disease so that we can stop it in earlier steps. For this, we first took the concept of data mining to find support; generated support is used as the weight of the symptom, which will be the initial pheromone value of ant. Then we consider pain in the chest, radiation on the back, feeling suffocated (heartburn), nausea, excessive weakness, and irregular heartbeat, a heart attack. This research work uses benchmark performance metrics i.e. sensitivity, specificity, and classification accuracy.