Hybrid Computational Intelligence Technique: Eczema Detection

This paper discusses the approach for detecting Eczema which can upsurge patients' risk of evolving allergic chaos like asthma, fever, and food allergy, in addition with other health conditions like obesity and cardiovascular disease, using computational intelligence techniques. We have proposed a novel deep learning framework approach which employs a pre-trained network integrated with AdaBoost algorithm for eczema detection. Its detection becomes cumbersome task due to the unavailability of a proper dataset, so a new dataset has been created. We claim that our proposed approach has outperformed the state-of-the-art algorithms (Naïve Bayes, K-Nearest Neighbors, Support Vector Machine and Neural Network) with an accuracy of 97.5%.