Journal article Open Access

A Machine Learning Approach for Heart Attack Prediction

Suraj Kumar Gupta; Aditya Shrivastava; Satya Prakash Upadhyay; Pawan Kumar Chaurasia

Dublin Core Export

<?xml version='1.0' encoding='utf-8'?>
<oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
<dc:contributor>Blue Eyes Intelligence Engineering  and Sciences Publication (BEIESP)</dc:contributor>
<dc:creator>Suraj Kumar Gupta</dc:creator>
<dc:creator>Aditya Shrivastava</dc:creator>
<dc:creator>Satya Prakash Upadhyay</dc:creator>
<dc:creator>Pawan Kumar Chaurasia</dc:creator>
<dc:date>2021-08-30</dc:date>
<dc:description>A heart attack also known as cardiac arrest, diversify various conditions impacting the heart and became one of the chief-reason for death worldwide over the last few decades. Approximately, 31% of total deaths globally are due to CVDs. It constitutes the pinnacle of chronic processes which involve complex interactions between risk factors which can and cannot be improved. Most of the instances or cases of cardiovascular diseases can be allocated to revisable risk factors where most of the instances are considered preventable. ML became the enhancing approach for the evolution of predictive models in health care industries and was decided to test various algorithms to check what extent their prediction scores estimate or ameliorate upon the results acquired. Researchers deploy various machine learning and data mining techniques over a set of enormous data of cardiovascular patients to attain the prediction for heart attacks before their occurrence for helping healthcare industries and professionals. This research comprises various Supervised ML classifiers like, Gradient Boosting, Decision Tree, Random Forest and Logistic Regression that have been used to deploy a model for Myocardial Infarction prediction. It uses the existing datasets from the Framingham database and others from the database of the UCI Heart repository. This research intends to ideate the prediction for probabilities of occurrence of a heart attack in the patients. These classifiers have been deployed in pipeline approach of machine learning to attain the prediction using both ways i.e., without optimizations and feature transformations as well as vice-versa. The results impersonate that the Gradient Boosting classifier is achieving the highest accuracy score in such a way that prediction used by our model is of binary form in where 1 means a chance of heart attack and 0 means no chance. Some of the most influential attributes are chest pain type among which the typical angina is the most influential and asymptotic chest pain is least, cholesterol level in which the level greater than 200mg/dl are more prone, increased heart rate, thal, and age. It is concluded that premature heart attack is preventable in 80% of the total cases just by using a healthy diet along with regular exercises and not using tobacco products also the person who drinks more than 5 glasses of water daily are less likely to develop attacks.</dc:description>
<dc:identifier>https://zenodo.org/record/5412827</dc:identifier>
<dc:identifier>10.35940/ijeat.F3043.0810621</dc:identifier>
<dc:identifier>oai:zenodo.org:5412827</dc:identifier>
<dc:language>eng</dc:language>
<dc:relation>issn:2249-8958</dc:relation>
<dc:rights>info:eu-repo/semantics/openAccess</dc:rights>
<dc:rights>https://creativecommons.org/licenses/by/4.0/legalcode</dc:rights>
<dc:source>International Journal of Engineering and Advanced Technology (IJEAT) 10(6) 124-134</dc:source>
<dc:subject>Cardiovascular-Disease, Framingham Model, Gradient Boosting, Machine Learning, Mayo-cardinal Infarction, UCI Model</dc:subject>
<dc:subject>ISSN</dc:subject>
<dc:subject>Retrieval Number</dc:subject>
<dc:title>A Machine Learning Approach for Heart Attack Prediction</dc:title>
<dc:type>info:eu-repo/semantics/article</dc:type>
<dc:type>publication-article</dc:type>
</oai_dc:dc>

28
30
views
downloads
 Views 28 Downloads 30 Data volume 26.2 MB Unique views 21 Unique downloads 27