<![CDATA[This study aims to classify skin cancer based on dermoscopic images using texture feature extraction through the Gray Level Co-occurrence Matrix (GLCM) technique by comparing the performance of four machine learning algorithms: Support Vector Machine (SVM), K-Nearest Neighbor (K-NN), Decision Tree, and Random Forest. This approach was developed to address the limitations of previous studies, which typically employed only a single algorithm without comprehensive comparison. The evaluation results show that Random Forest achieved the best performance, with an accuracy of 92.72%, precision of 94.44%, recall of 92.39%, and an F1-score of 93.40%. This is attributed to its ensemble nature, which combines multiple decision trees through a voting mechanism, making it effective in handling imbalanced data and complex texture patterns. Conversely, Support Vector Machine (SVM) demonstrated the lowest performance, with an accuracy of 66.06%, precision of 84.44%, recall of 64.40%, and an F1-score of 73.07%, indicating its limitations in recognizing nonlinear in high-dimensional data. Based on these results, the combination of GLCM and Random Forest has proven to be effective and optimal for medical image classification, and holds significant potential to support more accurate clinical decision-making in the early detection of skin cancer]]>
Copyrights © 2025
