Рақамли технологияларнинг назарий ва амалий масалалари 8-jild 2-son (2025) · 7-16-betlar
Comparative Analysis of Quantum Models in the Prognosis of Cardiovascular Diseases
Адилова, Ф.Т., Давронов, Р.Р.
Annotatsiya
This paper analyzes two approaches to the prediction of cardiovascular diseases, which have a binary classification problem based on the same Cleveland benchmark and aim to identify the advantages of quantum computing over traditional classifiers. This is a hybrid quantum random forest (HQNN, HQRF) for predicting the development of coronary heart disease in the early stages and an explainable prediction of cardiovascular diseases by the ensemble-quantum learning of Bagging-QSVC. The first approach uses various feature selection methods, which usually require a lot of computational effort, while the HQRF model and the earlier HQNN model are high-speed algorithms due to the nature of quantum computing. Numerical results show that HQRF is more suitable for small datasets, while HQNN is better suited for large datasets. In the second approach, the Bagging-QSVC model uses a quantum support vector classifier as a basic classifier. The results of the model are explained through the importance of the contribution of each individual feature using the Shaply (SHAP) algorithm. Comparative studies of other quantum classifiers on the Cleveland benchmark show the superiority of Bagging-QSVC with an accuracy of 90.16%. It follows from this that quantum machine learning classifiers are more effective than classical machine learning classifiers in predicting diseases of the cardiovascular system.
машинное обучениеансамблевое обучениеквантовое машинное обучениеобъяснимое машинное обучениепрогнозирование сердечных заболеванийmachine learningensemble learningquantum machine learningexplainable machine learningheart disease prediction
Metadata manbasi: jurnal OAI-PMH arxivi · Sindex toʻliq matnni saqlamaydi, manbaga havola beradi.