Kernel Parameter Tuning to Tweak the Performance of Classifiers for Identification of Heart Diseases

Kernel Parameter Tuning to Tweak the Performance of Classifiers for Identification of Heart Diseases

Annu Dhankhar, Sapna Juneja, Abhinav Juneja, Vikram Bali
Copyright: © 2021 |Pages: 16
DOI: 10.4018/IJEHMC.20210701.oa1
Article PDF Download
Open access articles are freely available for download

Abstract

Medical data analysis is being recognized as a field of enormous research possibilities due to the fact there is a huge amount of data available and prediction in initial stage may save patient lives with timely intervention. With machine learning, a particular algorithm may be created through which any disease may be predicted well in advance on the basis of its feature sets or its symptoms can be detected. With respect to this research work, heart disease will be predicted with support vector machine that falls under the category of supervised machine learning algorithm. The main idea of this study is to focus on the significance of parameter tuning to elevate the performance of classifier. The results achieved were then compared with normal classifier SVM before tuning the parameters. Results depict that the hyperparameters tuning enhances the performance of the model. Finally, results were calculated by using various validation metrics.
Article Preview
Top

1. Introduction

The primary reason of higher death rate around the world is linked with Cardiovascular diseases whether it is a developed, developing or under developed country (Ayatollahi et al., 2019). In order to have a better understanding of the conditions and diseases that impact the heart and its vessels, it must be explored that along with the circulatory system of blood may have some relationship to the coronary vascular diseases (Winker et al., 2015) (Task et al., 2013). Heart is one of the vital organs of the human body structure and furnishes blood to every corner of our anatomy. If heart discontinues to function as expected, then the other body parts like brain and multiple other organs will cease to function, and this can further cause sudden death of person. Heart diseases have emerged as one of the principal factors of deaths throughout the world. In the current competitive and busy lifestyle scenario, the detected cardiovascular diseases (CVD) are escalating on a daily basis. The World Health Organization (WHO) approximates that nearly 17 million people fail to survive every year pertaining to these cardiovascular disease, primarily caused due to heart attacks and strokes (Yamashita et al., 2018). Overweight, obesity, hypertension, hyperglycaemia, and high cholesterol are few potential drivers that are most eventful in triggering heart related issues. Additionally, American Heart Association (Stewart et al., 2019) cited symptoms including sudden gain in weight (1–2kg per day), disordered sleep pattern, swelling of body parts, cough and amplified heart beat rate (Parasuraman et al., 2019). That is the reason it became evident to annotate the extremely governing indicators and healthy lifestyles that can contribute to CVD. Prior to diagnosis of CVD there are multiple tests performed, which includes blood pressure, ECG, blood sugar, cholesterol etc. Typically such tests are generally repeated in case the patient becomes critical in condition and patient must begin taking medical aid immediately, it often becomes critical to give priority to tests (Rumsfeld et al., 2016). There are a countless heart diseases which includes failure of heart, stroke and coronary artery diseases as well (Jatav & Sharma, 2018). With respect to medical analysis, there has been a huge demand for identifying or predicting the diseases a person is potentially suffering from, before the start of damage by that disease, with only knowing about its symptoms. Working with a few classifiers and by analysing the symptoms we can predict the actual disease with which a person can suffer in near future. The presence of innovative technologies like machine learning, we can reasonably justify the matches that coexist among the data very quickly.

Complete Article List

Search this Journal:
Reset
Volume 15: 1 Issue (2024)
Volume 14: 1 Issue (2023)
Volume 13: 5 Issues (2022): 4 Released, 1 Forthcoming
Volume 12: 6 Issues (2021)
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing