Heart Failure Predictive Analysis Using Decision Tree Classification

Abstract

With an average age of 28 compared to Western countries, India's young population accounts for half of heart attacks in South Asia, which happen to people under 52. Autopsy reports, which identify the actual cause of death, frequently concentrate on sudden deaths in young adults that have no apparent reason or warning signs. Fat accumulation in the blood vessels of the heart is the cause of abrupt, unexpected natural deaths. The heart stops beating and loses blood as a result of these arteries narrowing or blocking. The body may exhibit subtle symptoms prior to abrupt death, such as shortness of breath, palpitations, tightness in the chest, and chest discomfort. A decision tree classification is a death dataset model that generates labelled classes at leaf nodes and makes judgments at edges to predict class labels for subsequent records. The purpose of the proposed paper study is to predict abrupt natural deaths, which are frequently brought on by smoking, by using regression analysis, a statistical technique that establishes the relationship between independent and dependent variables. The experiment's outcome, which looks at how Artificial Neural Networks (ANN) may be used to forecast heart failure, shows five records out of 50,000 patients from different hospitals. Perceptron’s, both single- and multi-layer, were used to gather patient information. Keywords: Artificial Neural Network, Linear Regression Analysis, Sudden and Unexpected Natural Deaths etc.