Using the Particle Swarm Optimization algorithm to design a Heart Disease Prediction System

• Main Authors: Chi-Hung Lai, 賴佶泓
• Other Authors: Yen-Chu Hung
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/8dcy56

碩士 === 國立嘉義大學 === 資訊工程學系研究所 === 105 === Heart disease is the leading cause of death in the near years. According to the reports of the American Heart Association and the American Stroke Association. Heart disease is the leading global cause of death. Heart disease accounts for 1 in 7 deaths in the U.S. To confirm the diagnosis of heart disease always takes much time and needs to perform multiple checks. The will lead to miss the heart disease discovered early and the heart disease treated early. For the correct diagnosis and the effective treatment of heart disease, a fast and accurate diagnosis system needs to be developed. This paper extracts the effective information by data mining, through the management of the discretization and standardization, and combines with the Modified Particle Swarm Optimization algorithm, the precious knowledge data can be obtained from medical information and health care system. These knowledge data can be applied to confirm heart disease early and let the patients accept effective treatment early. This study uses the University of California data set and combines with the algorithm to design a heart disease prediction system. First, the patients’ data will be transformed and converted into a convenient management form. Then, the data be divided into the heart disease group and the non-heart disease group. The Heart Disease Prediction algorithm can find the global best solution of the heart disease group and the global best solution of the non-heart disease group. To apply the two global best solutions can to predict whether the tested patient has heart disease or the tested patient has not heart disease. According to the study results, the designed Heart Disease Prediction system has a prediction accuracy more than 81%.