Yüksek performanslı makine öğrenme sistemi kullanılarak kardiyovasküler (CVD) hastalığının tanı

One of the most prevalent illnesses in middle-aged people is heart failure. Cardiovascular disease (CAD) is a common coronary disease with a high mortality risk of the various forms of heart disease. Medical imaging, such as angiography, is the most effective technique for diagnosing CAD. Angiography, on the other hand, is notorious for being pricey and causing a variety of side effects. In the first part of our work, we will address one of the most sophisticated and effective systems in the medical sector, specifically cardiac diagnostics, in this thesis. This advanced methodology, which is split into two sections, was developed with the aid of the best programming method, Python. To start, feature engineering and preprocessing with a high-performance re-sampling system and a neighborhood cleaning rule is used (NCL). Second, hyper-parameter optimization, API advanced features, and super learner models are examples of advanced and optimization machine learning models. We have used streamlined SVM to construct a high-performing diagnostic model. We built advanced architectures for high-performance machine learning, such as AutoML, advanced XGBoost, and advanced ensemble bagging models, in the second part of our study. We obtained the best performance using Auto-ML and advanced SMOTE method, which had an accuracy of 87%, advanced deep learning had an accuracy of 80%, and advanced bagging models had an accuracy of 82 percent.in addition, we achieved 86% with XGBoost and 92% used stacking model. Finally, we assume that our advances can change the way doctors view machine learning using advanced and high-performance machine learning tools, as well as promote the wider usage of Artificial Intelligence (AI) techniques in clinical settings. Keywords: CVD disease, NCL, Advanced machine learning models, Optimization, API function advanced, Super learner, Advanced Bagging models, Advanced XGBoost, AutoML, Python