AI Based Prediction for Heart Disease: A Comparative Analysis and an Improved Machine Learning Approach

• Published in: 2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT) pp. 1 - 9
• Main Authors: Raval, Jay, Verma, Jai Prakash, Islam, Sardar N. M., Jain, Rachna, Thakur, Narina
• Format: Conference Proceeding
• Language: English
• Published: IEEE 09.12.2022
• Subjects: Cardiac arrest; Classification algorithms; Data mining; Heart; Heart disease data set; Machine learning; Machine learning algorithms; Multi-Layer Perceptron (MLP); Prediction algorithms; Predictive analysis; Support vector machines
• DOI: 10.1109/ACAIT56212.2022.10137923

Heart disease problems are growing day by day in the world. Many factors are responsible for increasing the chance of heart attack and any other disease. Many countries have a low level of cardiovascular competence in predicting heart disease-related issues. Finding the best accurate machine learning classifiers for various diagnostic uses by data mining and machine learning techniques aids in predicting whether or not the heart disease-related issue will occur. To predict heart disease, a number of supervised machine-learning algorithms are used and their effectiveness are evaluated. With the exceptionof MLP and KNN, all applied algorithms had their estimated feature significance scores for each feature. This helps to find the main factors affecting heart disease and the accuracy of the model, which helps to get the best prediction. At the end of the research the support vector machine gives us 87.91 % highest testing accuracy compare with all applied machine learning algorithm.