Open AccessAutomated Detection of Coronary Artery Disease using Machine Learning Algorithm
Author(s) -
Dibakar Sinha,
Ashish Sharma
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1116/1/012151
Subject(s) - computer science , naive bayes classifier , machine learning , data collection , algorithm , coronary artery disease , data mining , artificial intelligence , statistical classification , medicine , support vector machine , statistics , mathematics
Data mining, an excellent development technology for discovering and gathering essential knowledge from vast data collection that can help analyze and draw up trends for decision-making in the industry. Talking about the medical sphere, data mining can be used to uncover and withdraw useful data and trends that can be helpful in clinical diagnostic results. The research focuses on the diagnosis of heart disease, taking past evidence and information into account. To achieve this SHDP, non-linear SVC with RBF kernel algorithms is designed to perfect this SHDP (Smart Heart Disease Prediction). The final is a useful algorithm to look for the right combination of hyper parameters to increase the precision of the algorithm (C, α). The requisite data was arranged in a structured way. The following features are derived from medical profiles for the estimation of the risks of heart failure in a patient: BP, age, sex, cholesterol, blood sugar, etc. The collected characteristics serve as an input to the Navies Bayesian heart disease prediction classification. The data collection used is divided into two parts, 80% of the data are used for preparation, and 20% are used for research. The method suggested includes data collection, user authentication, and log in (based on application), classification through Navies Bayesian, prediction, and safe data transmission via the AES application (Advanced Encryption Standard). An average accuracy, specificity, sensitivity, precision, 93.53% f-score, 89.22%, 91.24% and 86.98%, respectively. This method is also possible in clinical settings to help clinicians predict cardiac arrest.