<![CDATA[The incidence of breast cancer is steadily increasing each year in Indonesia. Currently, breast cancer is not only prevalent in the elderly population but also among younger individuals. Several studies indicate that breast cancer classification can be performed using ultrasonography. The aim of this research is to explore the use of the Random Forest method for classifying breast tumors in ultrasonography images based on texture characteristics. Although the model achieved 100% accuracy on the training data, testing with various folds showed a decrease in accuracy ranging from 51% to 54%, with varying precision and recall. Despite not being optimal, Random Forest demonstrates potential as a classification algorithm, providing a foundation for further development to enhance the accuracy of breast tumor diagnosis. Factors such as feature selection, dataset quality, and model parameters are crucial considerations for future research to support more accurate diagnoses by healthcare professionals.]]>
Copyrights © 2024
