Article Preview
Top1. Introduction
Heart disease kills more people than any other cause of death each year, accounting for 31% of all fatalities worldwide. According to the WHO (world health organization), heart disease is the “top killer of human health”(Li et al., 2022). The word “heart disease” refers to a wide range of cardiac diseases. HeartDisease (HD), which impairs the circulation of blood to the heart, is the most prevalent this type of heart illness in the United States. According to the American Heart Association, unhealthy conduct that leads to hypertension, obesity and overweight, hyperglycaemias, and high cholesterol increases the risk of heart disease.(Gárate-Escamila et al., 2020) myocarditis, arrhythmias, cardiomyopathy, Congestive heart failure, coronary heart disease, angina pectoris, and congenital heart disease, are different types of heart illnesses(Prasanna & Vijaya, 2022) Heart illness raises the risk of stroke, and cardiac muscle disease, heart arrhythmias, pericardial disease, and other significant health complications. Although the specific origin of heart disease is unknown, it has been proven that genetics, environment, and lifestyle might influence the probability of acquiring the condition (Mansourypoor & Asadi, 2017). The common symptoms are Shortness of breathing, Pain in other areas of the body, and irregular heartbeats, including the arms and left shoulder. heart attack can be caused by a reduction in blood flow. Various factors have been discovered as symptoms of heart failure and heart disease (HD) through medical and clinical studies. The risk factors can be divided into two categories: The first group of risk variables comprises those that cannot be changed, such as family history, gender, and age. The second group of risk factors includes things like cholesterol, smoking, and eating habits. It can be eliminated or controlled by modifying their lifestyle and taking medicine. (Javeed et al., 2019) By selecting the most relevant attributes, feature selection methods may be useful for lowering the cost of treatment (Abdollahi & Nouri-Moghaddam, 2022). Early identification, therapy, and healing are all aspects of heart disease. As a result, early detection of heart disease is essential for effective therapy. To learn about the patient's cardiovascular condition (Zhang et al., 2021). So early and correct diagnosis, as well as the delivery of suitable treatments, are critical to decreasing the number of deaths caused by heart disease. Such services must be available for those who are at elevated risk of heart disease. Several factors influence the heart disease forecast. Previously, authors concentrated much more on identifying important features using old methods like ANOVA, Chi-square, and ReliefF, backward feature selection with traditional classifiers in their heart disease prediction models, and emphasis was placed on defining the relationships between these features and reduced time complexity.
The increased healthcare data collection presents a new opportunity for physicians to improve patient treatment in the healthcare industry, with machine learning becoming an important solution to aid the diagnosis of patients.
In, heart disease prediction a huge variety of machine learning classification techniques which include K-Nearest Neighbour (KNN), Detection Tree (DT), Naïve Bayes (NB), Neural Network (NN), Support Vector Machine (SVM), and Logistic Regression (LR) play a key role for prediction of disease (Shankar et al., 2020). The identification of characteristics of cardiac illness by using computerized approaches like Artificial Neural Network (ANN), Multilayer perceptron (MLP), Deep Neural Network (DNN), Recurrent Neural Network (RNN), radial basis function (RBF), neural network classifier (NNC) and Deep Learning classifiers that may be anticipated for diagnostics has gotten a lot of attention. Normal and aberrant two-class classification procedures are available (Bharti et al., 2021).
This work aims to develop an evaluative heart disease forecast model for the healthcare industry. The forecast model is used to improve the accuracy rate, and true disease, decrease the miss-classification error, and more importantly, identify heart disease patients at an early stage. Predicting heart disease has two major issues: the first one is that the number of patients affected by heart disease and non-disease patients is imbalanced. The next issue is the number of irrelevant attributes of the available patient data set.