Seybold Report ISSN: 1533-9211
Ashish Kumar Gangwar, Priyansh Kamthan, Dr. Arvind Dagur
Vol 18, No 3 ( 2023 ) | Licensing: CC 4.0 | Pg no: 222-232 | Published on: 30-03-2023
Abstract
Heart disease, often known as Coronary artery disease, is the paramount source of death on earth over the past few years. It encompasses a variety of heart-related conditions. It is important to have timely access to reliable, practical, and accurate techniques for disease management and early detection in order to address a number of risk factors for heart disease. In the health maintenance manufacturing, data mining is a frequently utilised approach to handle vast amounts of data. Forecasting heart illness is necessary, researchers analyse vast amounts of complex medical data using a variation several approaches for data excavation and machine learning. The model in this project is based on supervised learning methods including Decision Tree, Decision Tree also Random Forest with K-nearest Neighbor along Support Vector Machine Classifier. It presents numerous heart disease-related features. It utilises the contempory dataset from the Cleveland database from the UCI Coronary artery disease patient history. The gathering has 1026 occurrences with 76 characteristics. Just 14 from the 76 qualities, which were essential to demonstrating how well dissimilar algorithms work, are used in the testing process. Predicting the chance that patients will acquire heart disease is the aim of this research project. According to the results, the Random Forest Classifier Algorithm has the highest accuracy rating.
Keywords:
Heart disease Prediction Using Decision Trees, Support Vector Classifiers, Logistic Regression, Random Forest, and KNN similar to excessive cholesterol, obesity, an increase in triglyceride levels, high blood pressure, etc., are to blame for the growth in the risk of heart disease. But, as time passes, There are many hospital patient records and examination data available. The patient's records can be retrieved from a variety of public sources, and further investigation can be done so that a variety of computer technologies can be used to properly analyse the patient's medical data that pinpoint the illness in order to prevent it from becoming fatal.