Article Preview
Top1. Introduction
The oxygen-rich blood is pumped to the full body constantly through the network of arteries and veins, which dignify the predominant responsibility of the heart, and any interruption caused to this process can be termed coronary disease (Rani, et al., 2021). Coronary heart disease is a deadly disorder, which makes the life of the individual fatal (Sarath, 2017; Banu & Gomathy, 2014; Krishnaiah, 2014). The researchers highlight that the people affected with heart disease, particularly heart attacks, fail to survive. Such disease has occurred in the male of middle age or old age people when contrasted with females, and the augmenting rate of cardiovascular diseases leads to higher mortality rates, which in turn causes a significant burden to the worldwide healthcare systems (Shaji, 2019; Trevisan, et al., 2020; Yadav & Pal, 2020). The delighted fact is that the possibility of an occurring heart attack on children will be comparatively lesser than the older adults https://www.ahajournals.org/doi/full/10.1161/01.cir.99.9.1165 (Subhadra & Vikas, 2019; Ghosh, et al., 2021). Heart diseases can be classified into various types: videlicet coronary heart disease, otherwise called Atherosclerosis, congenital heart anomaly, arrhythmia is also known as cardiac dysrhythmia, and so on and the patient agonizing from this deadly disease have copious symptoms, the symptoms are as follows chest pain, deep sweating, staggering sensations and the major cause behind the heart disease is occurred due to smoking, high blood pressure, diabetes, obesity and various other reasons (Razmjooy, et al., 2018). The diagnosis of the disease is the time consuming process, when it is carried out manually (Bhambere, 2017). Protruding methods of diagnosing heart disease are immoderate and excruciating, which creates a necessity for the non-invasive way of diagnosing heart disease at a reasonable cost (Rani, et al., 2021).
Recent accomplishments in the field of medical field stimulate the necessity of Machine Learning systems for health monitoring applications (Magesh & Swarnalatha, 2021) (Balakrishnan, et al., 2021; Juneja, et al., 2021a). A decision support system is designed utilizing the machine learning techniques and clinical data for detecting heart disease, and into the bargain, the cost is also reduced. The detection of heart disease accurately, the distinct decision should be made by designing the computational tools (Reddy & Khare, 2018; Gadekallu & Gao, 2021; Reddy & Khare, 2017). The tools mentioned above believe in optimization (Moallem, et al., 2013), clustering, deep learning (Obulesu, et al., 2021) (Juneja, et al., 2021b), and Machine Learning computational methods (Mousavi, et al., 2011; Amin, et al., 2019; Magesh & Swarnalatha, 2021). The researchers make the predictions accurately and precisely through the large databases available in the repositories. The mortality rate related to CVDs, the recent studies are focused on heart-related issues in adults and children (Ghosh, et al., 2021). Machine learning algorithms mostly depend on the unchangeable characteristics of the training and testing data. Hence, for more accurate prediction, the feature selection methods, namely data mining, relief selection, and Least Absolute Shrinkage and Selection Operator (LASSO), can help manipulate the data. It emphasizes that feature selection is more effective when trained with suitable datasets (Singh & Samagh, 2020; Mienye, et al., 2020; Wang, et al., 2020; Tama, et al., 2020). The chances of disease occurrence can be predicted by selecting the more suitable features, classifiers, and hybrid models (Mishra & Tarar, 2020; Kausar, et al., 2016) (Sampathkumar, et al., 2020). The lack of depth analysis due to the unavailability of medical datasets, irrelevant feature selection, and complexity in ML algorithm applications, the lack of depth analysis stands as a pitfall for the accurate detection of heart disease (Ghosh, et al., 2021).