<![CDATA[Heart disease is a leading cause of mortality globally, frequently progressing without overtly visible symptoms during its onset; thus, a data-intensive solution is warranted to support early screening. This paper proposes an analysis of heart disease risk factors using clinical data through the application of descriptive statistics, correlation analysis, and logistic regression. The clinical dataset utilized is a secondary compilation containing information on 918 patients, featuring 12 numerical and categorical variables alongside one dependent variable: heart disease. Descriptive statistics were employed to elucidate the characteristics of the clinical data, while Pearson correlation analysis and Chi-Square tests were utilized to examine associations among the clinical parameters. Logistic regression analysis served as the primary methodology for determining heart disease risk. The findings indicated that among the 918 patients, 55.3% were diagnosed with heart disease, with a peak incidence observed in middle-aged individuals. Pearson correlation analysis demonstrated no strong correlations between individual numerical variables and heart disease; however, categorical variables such as ChestPainType, ExerciseAngina, and ST_Slope were significantly associated with the condition. The logistic regression model achieved an accuracy of 88.59% and a recall value of 0.93 for the heart disease class. This paper underscores that an interpretive statistical approach can substantially support the development of clinical decision support systems for the early screening of cardiac patients.]]>
Copyrights © 2025
