Comparative Analysis of Skin Cancer Detection Using Classification Algorithms

The increment in death rates from skin cancer is due to the lack of awareness about the symptoms and its possible preventions. Owing to which detection at an early stage is complex in order to prevent cancer from spreading. This paper presents an automated and a precise approach for distinguishing between benign skin pigmented lesions and malignant melanoma. This detection system does not require special imaging equipment or conditions and may be used by the general population. The (DIP) digital image processing algorithms used for classification are as follows. First, the image data is pre-processed to remove noise, and adjust the dimensions and contrast with PCA dimensionality reduction. Then, the pixel values are normalized and scaled to improve the results. Feature selection is used to choose the best features. Numerous classifiers are employed (K-NN, SVM etc) to forecast the projected work's stages and efficiency. The recommended method is evaluated using well-known quantities, sensitivity, precision, accuracy, and specificity, with higher values achieved. The accuracy rate and r2 score is varied for each classifier algorithm when compared to existing methods with (ME) Model Evaluation Table.