International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 07 | July 2024
www.irjet.net
p-ISSN: 2395-0072
Heart Disease Identification Methods using Machine Learning and efficient data balancing techniques Ranu Masram1, Prof. Sudeep Kishore Sharma2, Mr.Nagendra Kumar3 1Reseacrh Scholar, Department of CSE, Shri ram Institute of Science and Technology, Jabalpur, M.P. 2Professor, Departmet of CSE, Shri Ram Institute of Science and Technology, Jabalpur, M.P.
3Professor, Departmet of CSE, Shri Ram Institute of Science and Technology, Jabalpur, M.P. ---------------------------------------------------------------------***-------------------------------------------------------------------doctors need to approach the diagnosis and prognosis of Abstract –A heart attack is a life-threatening event that is heart failure more specifically. New techniques for data very difficult to predict. Early diagnosis and prompt analysis can lead to early diagnosis of cvd through evaluation treatment can reduce mortality. According to the British of patients' medical records [2]. Forest and finally diagnosis. health foundation (b.h.f.), 1 In 14 people worldwide have The testing process will take longer than expected heart disease or heart disease. It is also estimated that depending on people's health and well-being, natural and approximately 200 million people suffer from heart disease. genetic factors in lifestyle. However, now healthcare can be Medical records consist of many different sections collected done by estimating and measuring risk for more serious from different sources, each offering a different perspective prevention of infection, thus achieving results resulting in on the patient's condition. Machine learning has proven to be better health (4). This has led to the development of an excellent method for predicting non-standard data. advanced techniques for analyzing heart disease (hd) clinical Algorithms such as svc, k nearest neighbor, decision tree, data to identify advanced failures. Many studies have used random forest classifier, XGBoost and cnn can be used for machine learning (ml) algorithms to predict cvd from clinical early detection of viruses. Data mining techniques are used data. However, medical records are still highly problematic to collect data from clinical sites and use it to determine due to class inequality and size. Therefore, using machine initial diagnoses of diseases without the need for learning without solving these problems reduces the intervention from doctors. Application form. Machine efficiency and accuracy of the method. Previous researchers learning algorithms are used especially to predict the risk of focused on specific selection (fs) and used various machine heart attack. This article is about using machine learning to learning methods to predict cvd. Uzun et al. [5] Developed a predict a patient's risk of heart attack. This research paper heart disease decision system based on rough system (rs) focuses on classification techniques in machine learning in and chaos firefly algorithm (cfars-ar) to select the best healthcare with the aim of finding similar models that can features, and then obtained a type 2 fuzzy logic system for make important predictions and help in early detection of HD detection. The design achieved an accuracy rate of diseases. Data on various characteristics such as age, gender 88.3%. Authors in [6] combined rs with Backpropagation and cholesterol levels were used to develop the prediction neural network (bpnn) to predict cvd. Additionally, Dwivedi models. The model was trained and tested on various data performed the comparison of hd prediction using different sets to determine its accuracy and ability to predict heart ml models such as artificial neural network (ann), logistic disease. The results of this study can be used to develop regression (lr), classification tree and naive bayes (nb). The more accurate methods to predict heart disease risk and authors concluded that lr outperformed other methods in reduce the number of deaths from heart disease. detecting cvd [7]. Additionally, huck et al. A comparative study was conducted using different ml models (e.g. Ann, rf Keywords: Heart attack prediction, Data balancing, Data and lr) with different fs (e.g. Click). The authors report that Pre-processing, Machine learnings, SMOTE, DBSMOTE, CNN. removing features affects the performance of the model. This 1. INTRODUCTION study concluded that lr with click fs achieved an accuracy of 89% compared to other methods used in the same study [8]. Cardiovascular disease, also known as cardiovascular Amin et al. Comparative analysis, analyzing the main disease (cvd), is the leading cause of death worldwide. In a features and using seven ml models including ann, lr, recent study, the world heart federation found that one-third decision tree (dt), support vector machine (svm), nearest of deaths was caused by heart disease [1]. According to neighbor (knn), nb and hybrid model (voting with lr and statistics from the world health organization (who), more voting). Was carried out. Note) [9]. The results showed that than 23.6 million people may die from cvd, mainly stroke the hybrid model achieved the best accuracy (87.41%). In and heart failure, by 2030 [2]. Cvd can be caused by many another study, a hybrid model consisting of rf and linear factors, including stress, alcohol, smoking, poor diet, poor model (hrflm) was developed to improve the accuracy of cvd lifestyle, and other health problems such as high blood prediction. This work has been applied to different variants pressure or diabetes. However, most cvd diseases are of the traits; later, vijayashree et al. A recent project completely treatable when diagnosed early [3]. In this case, introduced a weighting decision based on population
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 377