Hybrid Model of Machine Learning Algorithms for Prediction of Cardiovascular Disease

Main Article Content

Urja Desai, Shamla Mantri

Abstract

The utmost prevalent reason of death in worldwide is coronary artery disease, which is serious health issue for many people accounting for nearly 31% of all fatalities. Patients with common illnesses and symptoms are being reported in greater numbers. Many patients' lives can be saved by early identification of cardiovascular disorders such heart attacks and cardiovascular disease. It relates a slew of heart disease risk factors to the essential need for accurate, trustworthy, and practical approaches for early detection and management. In the healthcare, two popular approaches for evaluating huge datasets are mainly machine learning and deep learning algorithms. Many experts and researchers have practices a range of both deep learning and machine learning techniques for examine large quantities of complex medical data, assisting doctors in predicting heart problems. Accurately predicting, if a patient has cardiovascular diseases or not is the aim of this research. In this paper we have studied and implemented many traditional machine learning algorithms which are Naïve Bayes (NB), Logistic Regression (LR), Decision Tree (DT), Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boost (XGBoost) and K-Nearest Neighbor algorithm (KNN). Based on output we have implemented hybrid model to archive more accuracy. ROC-curve, Accuracy, Error rate, Recall, Precision and F1 score performance evaluation metrics which are applied to compare the effectiveness and performance of the methodologies in study. We have obtained 93.4% accuracy using this proposed hybrid model using a stacking classifier technique.

Article Details

Section
Articles