Skin Disease Classification using Convolutional Neural Network

Abstract

Due to the need for early diagnosis of certain disorders, the classification of dermatoscopic images has become an absorbing area of study. Due to its intricacy, dermatology is one of the most unexpected and challenging fields to diagnose. Based on the previous findings and results that were found, the custom model developed is more efficient than the other neural networks utilizing the dermnet for skin lesions (dataset). The performance of the system is improved by using data augmentation a more effective classifier. This study has the potential to make a difference to aid dermatologists in making more precise decisions in the clinic.