МАШИННОЕ ОБУЧЕНИЕ ДЛЯ КОМПЛЕКСНОЙ ОЦЕНКИ РИСКА СЕРДЕЧНО-СОСУДИСТЫХ ЗАБОЛЕВАНИЙ И БИОХИМИЧЕСКИХ ИЗМЕНЕНИЙ: АКЦЕНТ НА АСПАРТАТАМИНОТРАНСФЕРАЗЕ

Abstract

. This study explores various machine learning techniques for analyzing risk factors associated with cardiovascular diseases. Two approaches were employed to develop predictive models — XGBoost and a Convolutional Neural Network (CNN). The primary focus was to evaluate the performance of each model in classification and regression tasks, as well as their ability to identify key biomarkers and risk factors such as cholesterol, ferritin, homocysteine, and aspartate aminotransferase (AST). The XGBoost algorithm was fine-tuned for handling tabular data and demonstrated high accuracy in risk prediction. In contrast, the CNN model, while showing an initial reduction in training error, exhibited signs of overfitting during validation. A comparison based on metrics such as Mean Squared Error (MSE), Coefficient of Determination (R²), Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC) revealed significant differences between the two models. The findings confirm the efficiency of XGBoost in processing structured data and summarizing risk factor insights, while the CNN model requires further optimization to manage sparse datasets. Overall, the study highlights the importance of selecting an appropriate model architecture and tuning parameters to achieve reliable cardiovascular disease diagnosis.

Key words: cardiovascular diseases, machine learning technologies, Mean Squared Error, biochemical indicators, XGBoost, Vanilla CNN, Pathology