An optimal method for diagnosing heart disease using combination of grasshopper evalutionary algorithm and support vector machines

• Published in: Heliyon Vol. 10; no. 9; p. e30363
• Main Authors: Zhou, Wei, Liu, Hongbo, Zhou, Rui, Li, Jiafu, Ahmadi, Sina
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 15.05.2024; Elsevier
• Subjects: data collection; Data mining; death; decision making; disease diagnosis; grasshoppers; Heart disease; heart diseases; Locust evolutionary algorithm; medicine; Performance criteria; society; Support vector machine; support vector machines; Support vector machine; Data mining; Locust evolutionary algorithm; Heart disease; Performance criteria
• ISSN: 2405-8440; 2405-8440
• DOI: 10.1016/j.heliyon.2024.e30363

Due to the importance of accurate diagnosis and prompt treatment of this condition, the medical world is searching for a solution for its early detection and efficient treatment. Heart disease is one of the leading causes of death in modern society. With the development of computer science today, this issue can be resolved using computers. Data mining is one of the solutions for diagnosing this illness. One of the cutting-edge disciplines, data mining, can aid in better decision-making in many areas of medicine, including disease diagnosis and treatment. In order to improve diagnosis accuracy, a combination method using the evolutionary algorithms locust and support vector machine has been tested in this study. Use should be made of heart disease. Because of the hybrid nature of this approach, normalization is actually carried out in three steps: first, by using pre-processing operations to remove unknown and outlier data from the data set; second, by using the locust evolutionary algorithm to choose the best features from the available features; and third, by classifying the data set using a support vector machine. The accuracy criterion for the proposed method compared to Niobizin methods, neural networks, and J48 trees improved by 18 %, 30 %, and 24 %, respectively, after implementing it on the data set and comparing it with other algorithms used in the field of heart disease diagnosis.