DIAGNOSIS OF CARDIOVASCULAR DISEASE USING MACHINE LEARNING ALGORITHMS.pptx

DIAGNOSIS OF CARDIOVASCULAR DISEASE USING MACHINE LEARNING ALGORITHMS.pptx

1. DIAGNOSIS OF CARDIOVASCULAR DISEASE USING MACHINE LEARNING ALGORITHMS ASMA MALICA J - 21CSR011 DINESH K - 21CSR044 ELAVARASU P - 21CSR046

2. PROBLEM DEFINITION Heart disease is the global health issue and is responsible for considerable proportion of death worldwide Early detection and classification of heart disease in patients Develop methods to ensure data quality and handle missing or noisy data effectively. Explore feature selection and extraction techniques to improve model performance and interpretability. Address class imbalance through techniques like oversampling, undersampling, or using appropriate evaluation metrics.

3. INTRODUCTION Cardiovascular disease (CVD) remains a leading cause of morbidity and mortality worldwide, posing a significant health challenge that transcends geographic boundaries and demographic characteristics. Traditional diagnostic methods, though valuable, are not without their limitations. Clinical assessment and expertise, blood tests, electrocardiograms (ECGs), and medical imaging provide essential information but often require substantial time, are subject to human interpretation, and may fall short in detecting early-stage disease. Moreover, with the evolving landscape of healthcare, there is an ever-growing need for scalable and efficient diagnostic solutions that can be integrated seamlessly into clinical workflows. In this context, the integration of healthcare, especially machine learning offers an exciting opportunity To significantly transform the way we diagnose cardiovascular disease (CVD)

4. Machine learning algorithms can do a great job with large and complicated health data. They find important patterns and use the data to make predictions. These algorithms have the power to make CVD diagnosis more accurate and faster, which could mean doctors can help patients sooner and improve their health. This research is all about using machine learning to improve how we detect cardiovascular disease and tackle the many challenges that come with it. By using different types of healthcare data, advanced computer programs, and the knowledge of healthcare professionals, this study wants to make existing ways of diagnosing diseases better. It also aims to create new methods that can change and grow as healthcare evolves. In this paper, we will discuss our research goals, how we are doing the research, and what we have found. We want to use machine learning (like computer programs that learn from data) to help improve people's health. Our goal is to make it easier and faster to detect heart problems. This way, we hope to make people healthier and happier all over the world.

5. EXISTING WORK The proposal of unsupervised learning and clustering technique to categorize the Heart Disease Dataset. The variety of models such as decision tree, logistic regression, random forests, native Bayes and vector machines are used for the identification of heart illness diagnosis. The nine machine learning classifiers are used in the final datasets such as AB, LR, ET, MNB, CART, SVM, LDA,RF AND XGB. Data processing methods including regression and classification are used in cardiovascular dataset. Decision tree algorithm is used to find the final output for the cardiovascular disease. The language used is python, Jupyter notebook of the Anaconda Navigator to run a python code efficiently compared to python IDE and Microsoft V isual. The Datasets are taken from IEEE DataPort and Kaggle.

6. PROPOSAL WORK The ten machine learning classifiers are used in the final datasets such as AB, LR, ET, MNB, CART, SVM, LDA,RF AND XGB, HYBRID-SVM. The final output will be decided by the hybrid svm machine learning algorithm. The language used is python, Jupyter notebook of the Anaconda Navigator to run a python code faster and efficient than other compiler. SVMs are known for their strong classification abilities, especially when dealing with complex and high-dimensional data. The combining of Hybrid-SVM with other techniques will produce a better performance The gaining output from the Hybrid-Svm will be more accurate than the output from other techniques.

7. DATASET