Sand Cat Swarm Optimization with Deep Transfer Learning for Skin Cancer Classification

• DOI: 10.32604/csse.2023.038322
• Publication Date: 2023-07-26T11:34:57Z
• Authors: C. S. S. Anupama; Saud Yonbawi; G. Jose Moses; E. Laxmi Lydia; Seifedine Kadry; Jungeun Kim
• Keywords: Benchmark (surveying); Feature (linguistics); Skin lesion; Transfer of learning; Cancer Detection

Skin cancer is one of the most dangerous cancer. Because high melanoma death rate, skin divided into non-melanoma and melanoma. The dermatologist finds it difficult to identify from dermoscopy images lesions. Sometimes, pathology biopsy examinations are required for diagnosis. Earlier studies have formulated computer-based systems detecting lesion images. With recent advancements in hardware software technologies, deep learning (DL) has developed as a potential technique feature learning. Therefore, this study develops new sand cat swarm optimization with transfer method detection classification (SCSODTL-SCC) technique. major intention SCSODTL-SCC model lies recognition different types on dermoscopic Primarily, Dull razor approach-related hair removal median filtering-based noise elimination performed. Moreover, U2Net segmentation approach employed infected regions Furthermore, NASNetLarge-based extractor hybrid belief network (DBN) used classification. Finally, performance can be improved by SCSO algorithm hyperparameter tuning process, showing novelty work. simulation values scrutinized benchmark dataset. comparative results assured that had shown maximum measures.