Machine Learning and Ensemble Approach Onto Predicting Heart Disease

TL;DR

This paper explores using multiple machine learning models and ensemble techniques to improve the accuracy of heart disease diagnosis from clinical data.

Contribution

It introduces an ensemble approach combining various classifiers to enhance predictive performance for heart disease detection.

Findings

Ensemble method outperforms individual classifiers in accuracy.

Support Vector Machine and Random Forest show high individual performance.

The approach facilitates early and accurate diagnosis of heart disease.

Abstract

The four essential chambers of one's heart that lie in the thoracic cavity are crucial for one's survival, yet ironically prove to be the most vulnerable. Cardiovascular disease (CVD) also commonly referred to as heart disease has steadily grown to the leading cause of death amongst humans over the past few decades. Taking this concerning statistic into consideration, it is evident that patients suffering from CVDs need a quick and correct diagnosis in order to facilitate early treatment to lessen the chances of fatality. This paper attempts to utilize the data provided to train classification models such as Logistic Regression, K Nearest Neighbors, Support Vector Machine, Decision Tree, Gaussian Naive Bayes, Random Forest, and Multi-Layer Perceptron (Artificial Neural Network) and eventually using a soft voting ensemble technique in order to attain as many correct diagnoses as possible.