Skin Lesion Classification In Dermoscopic Images With High Accuracy Using K-Nearest Neighbor Algorithm Comparing With Random Forest Method

The purpose of this research is to evaluate the efficacy of the K-Nearest Neighbor algorithm in comparison to the Random Forest approach in terms of the accuracy and specificity of skin lesion categorization in dermoscopic pictures. Both the Materials and the Methods: The techniques of K-Nearest Neighbor (KNN) and Random Forest were utilised in order to classify forty samples of dermoscopic pictures including skin lesions as either benign or malignant. The technique of identifying skin lesions typically involves four stages: pre-processing, segmentation, feature extraction, and classification. The end result was that the KNN acquired an accuracy of 93.27 percent and a specificity of 89.10 percent, in comparison to the accuracy of 88.56% and the specificity of 85.38% obtained by the Random Forest. Since p is less than 0.05, the differences between the two groups are statistically significant. According to this study's findings, the accuracy achieved in detecting skin lesions using KNN classifiers is superior to that achieved using Random Forest classifiers. [Citation needed] [Citation needed]