تفاصيل الرسالة العلمية

Improving the clustering of medical data using the balanced iterative reducing and clustering (BIRCH) hierarchical clustering algorithm increases the patients' diagnostic process accuracy. As a result, patients' medical services are improved, and there is a noticeable enhancement of the processing efforts and time. This thesis's main objective is to increase the accuracy of the BIRCH clustering algorithm by proposing an improvement over the hierarchical BIRCH cluster algorithm when applied to medical data. The proposed enhancement is based on combining three steps with the original BIRCH algorithm. These steps are: applying the scaling algorithm, transforming data, and using a modified version of the Euclidean distance. The used Euclidean distance equation is enhanced from the original equation used in the original BIRCH algorithm. The improved BIRCH algorithm was tested using a medical dataset for breast cancer diagnosis. Finally, the proposed modified algorithm was compared with the original BIRCH algorithm. The experimental results showed that the modified algorithm has a higher accuracy rate than the original, as the accuracy of the modified algorithm has reached 90%. Compared to 22% for the original algorithm