Detection and classification of Melanoma image of skin cancer based on Convolutional Neural Network and comparison with Coactive Neuro Fuzzy Inference System

Abstract

The main aim of the work is to improve the accuracy for the skin cancer detection that leads to identification of skin disease in a preclinical stage using Convolutional Neural Network algorithm in comparison with Coactive Neuro Fuzzy Inference System. The datas are collected from the open access website uci machine learning repository datasets. In detection of skin disease, 20 Melanoma images (MI) are used for Convolutional Neural Network (Group 1) and it is compared with Coactive Neuro Fuzzy Inference System (Group 2) with 80 % pretest power and maximum accepted error as 0.05. Proposed system using CNN improves accuracy to 98.31 % compared with CANFIS with an accuracy of 87.61%. Significance value is 0.001 (p i 0.05, 2-tailed). In this view the detection of skin cancer using CNN is better compared to CANFIS.