International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 01 | Jan 2024
p-ISSN: 2395-0072
www.irjet.net
CARDIOVASDCULAR DISEASE PREDICTION USING MACHINE LEARNING Tadisina Hasini1, B. Akhileshwar, D. Prakash3, Dr. D. Sreenivasulu4 1B-Tech 4th year, Dept. of CSE(DS), Institute of Aeronautical Engineering
2 B-Tech 4th year, Dept. of CSE(DS), Institute of Aeronautical Engineering 3 B-Tech 4th year, Dept. of CSE(DS), Institute of Aeronautical Engineering
4Associate Professor, Dept. of CSE(DS), Institute of Aeronautical Engineering, Telangana, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Cardiovascular diseases are a predominant cause
quality data and optimal model performance. In order to increase the model's potential to predict outcomes consistently across all risk categories, the Synthetic Minority Over-sampling Technique (SMOTE) is applied to correct the inherent imbalance in class distribution.
of mortality worldwide, highlighting the necessity for early detection and efficient risk management.. This work uses a publicly available cardiovascular disease dataset and the Multinomial Naive Bayes method to estimate the risk levels for CVD. Thorough preparation was performed on the data, which included discretizing variables, managing missing values, and normalizing continuous characteristics. To rectify the class imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) was utilized. The target variable was classified as either disease-related or not. Grid Search Cross-Validation was used for hyperparameter adjustment in order to maximize model performance. When tested on a test set, the finished model showed good classification performance and acceptable accuracy. This study demonstrates how machine learning may be used to forecast the risk of cardiovascular disease, offering a useful tool for early detection and preventative healthcare measures.
The Multinomial Naive Bayes technique, recognized for its simplicity and efficacy with categorical data, is utilized here to assess a large dataset combining demographic information and clinical characteristics. These include age, cholesterol levels, blood pressure measurements, and other key health markers. By processing and analyzing these datasets, the research seeks to construct a robust prediction model capable of predicting existence of CVD’s.
1.1 Existing System Supervised Learning Algorithms: Commonly used algorithms include logistic regression, decision trees, random forests, support vector machines (SVM. These algorithms learn from labeled data and are used for tasks such as classification (e.g., predicting disease risk categories) and regression (e.g., predicting continuous outcomes like blood pressure).
Key Words: Cardiovascular Disease Prediction, Multinomial Naive Bayes, Risk, Assessment, SMOTE, Hyperparameter Tuning, Grid Search CV, Health Informatics
Deep Learning: Deep learning techniques, such as neural networks and convolutional neural networks (CNNs), excel at extracting intricate patterns from large, unstructured datasets like medical images (e.g., MRI scans, CT scans) and electrocardiograms (ECGs). They are increasingly used for tasks like image classification and anomaly detection.
1.INTRODUCTION Cardiovascular diseases (CVDs) are a major cause of death and morbidity in many populations, making them a serious global health concern. These illnesses, which comprise a variety of heart and blood vessel problems, include heart failure, hypertension, coronary artery disease, and stroke. The Multinomial Naive Bayes method is the specific tool used in this study to forecast the risk levels associated with CVDs by utilizing machine learning techniques.
Unsupervised Learning: Algorithms like clustering and dimensionality reduction (e.g., principal component analysis) are used to explore and uncover hidden patterns and structures within data, potentially identifying new risk factors or patient subgroups in cardiovascular disease.
Advances in machine learning and computational techniques have recently surfaced as potential instruments in the healthcare industry, providing fresh strategies to improve patient care, diagnostics, and illness prediction. An estimated 17.9 million deaths were attributed to CVDs in 2019 alone, according to the World Health Organization (WHO), highlighting the critical need for efficient preventative and early intervention programs (WHO, 2020).
1.2 Proposed System This project proposes leveraging the Multinomial Naive Bayes algorithm for cardiovascular disease (CVD) prediction, addressing current limitations in existing machine learning approaches. The system will utilize a comprehensive dataset including demographic details and clinical parameters like age, cholesterol levels, and blood pressure. Data preprocessing will involve handling missing values, normalizing features, and discretizing continuous variables.
The dataset is meticulously preprocessed utilizing methods such resolving missing values, normalizing continuous features, and encoding categorical features to ensure high-
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 43