International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 08 | Aug 2023
p-ISSN: 2395-0072
www.irjet.net
Heart Failure Prediction using Different Machine Learning Techniques Prof. Pritesh Patil1, Rohit Bharmal2, Shravani Ghadge3, Dhanashri Gundal4, Ankita Kawade5 1Prof. Information Technology, AISSMS Institution of Information Technology, Pune, Maharashtra, India 2Student, Information Technology, AISSMS Institution of Information Technology, Pune, Maharashtra, India 3Student, Information Technology, AISSMS Institution of Information Technology, Pune, Maharashtra, India 4Student, Information Technology, AISSMS Institution of Information Technology, Pune, Maharashtra, India 5Student, Information Technology, AISSMS Institution of Information Technology, Pune, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - A This study compares the effectiveness of four
Future times may see an increase in their frequency. Heart failure might eventually result in death if people did not pay attention to it. The patient's medical and family histories, a physical examination, and test findings are often the foundation of the diagnostic process for heart failure. Due to several risk factors, including diabetes, high blood pressure, high cholesterol, an irregular pulse rate, and many other conditions, it can be challenging to diagnose heart disease. cardiac failure is a severe symptom or advanced stage of several cardiac disorders. Typically, cardiac ejection would be insufficient in patients with heart failure.
well-known machine learning methods for predicting heart failure using a publically accessible dataset from kaggle.com: Random Forest (RF), K-nearest neighbors (KNN), Naive Bayes (NB), and Logistic Regression (LR). LR, or regression. They were chosen as a result of their successful applications in the field of medicine. The care of heart failure situations has to be improved in order to raise the survival rate because it is a widespread public health issue. The availability of sophisticated computational systems and the abundance of medical data on heart failure allow researchers to carry out more tests. Accuracy, precision, recall, f1-score, sensitivity, and specificity were used to evaluate the effectiveness of the machine learning algorithms in predicting heart failure using 14 symptoms or characteristics. In comparison to KNN, NB, and LR, experimental investigation revealed that RF delivers the greatest performance score (90.16). The findings of additional RF trials to identify the key indicators of heart failure prediction showed that each of the 14 symptoms or traits is crucial.
Heart failure has a high mortality rate and is expensive to treat. Since heart disease is the most prevalent, it is urgent to develop very accurate and early methods of diagnosing heart disease, which can help many patients survive. There are several scanners available to identify heart illness, however detecting a cardiac ailment before it manifests itself can save many lives. By utilizing a tool that enables the administrator to visually assess the patient's data, we are giving further information to the administrator. Early detection and analysis of the existence of arrhythmia is crucial to preventing patients from developing heart problems. In many circumstances, the existence of a stroke or heart failure may be caused by the small levels of cardiac rhythm. The healthcare sector has a lot of promise with data mining since it can help health systems assess and diagnose diseases by using the data. The cost and time savings result from our ability to evaluate data and forecast illness.
Key Words: Machine Learning, Heart Failure Prediction, Logistic Regression, Naive Bayes, Random Forest, Knearest neighbors
1. INTRODUCTION The human body's most vital organ is the heart. The effective functioning of the heart is absolutely essential to human life. The heart delivers blood through blood vessels to the various bodily areas, with enough oxygen and other necessary nutritional elements for the organism's efficient operation. A healthy heart leads to a healthy life. But in the modern world, heart disease has emerged as a major factor in both male and female fatalities. cardiac failure results from corona virus induced cardiac muscle inflammation. Regardless of respiratory symptoms, experimental data indicates that 1 in 5 patients had cardiac damage caused by the Corona virus. The most prevalent kind of cardiac illness is coronary artery disease. Heart failure is a major issue that significantly affects people's lives. Most individuals consistently disregard their health due to the faster pace of life, larger portion sizes, and inactivity. Additionally, because of the deterioration of the environment, those factors may contribute to the problem of heart failure, which
© 2023, IRJET
|
Impact Factor value: 8.226
The World Health Organization predicts that heart fragility will cause the deaths of almost 23.6 million people between now and 2030. Therefore, anticipating a coronary illness should be avoided in order to lower the risk. There are two primary categories for heart disease risk factors. We cannot modify the risk variables in the first category, which includes things like age, gender, and family history. Risk factors in the second category include things like smoking, poor eating habits, and high cholesterol; we improve this second group. Therefore, by using the medical data mining classification algorithms, which is a crucial part for identifying the possibility of heart attack, the risk factors belonging to the second class can be eliminated or controlled by changing lifestyle and through medication. Medical professionals most frequently employ angiography
|
ISO 9001:2008 Certified Journal
|
Page 910