International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1644 Survey on Features and Techniques used for Skin Disease Classification Kamalapurada Vishwanath 1 , B Raghavendra Setty 2 , Puneeth G J 3 , Dr. H. Girisha 4 1, 2 P.G. Scholar, Department of Computer Science & Engineering, RYMEC, Ballari-583104, India 3 Asst. Professor, Department of Computer Science & Engineering, RYMEC, Ballari-583104, India 4 Professor, Department of Computer Science & Engineering, RYMEC, Ballari-583104, India ----------------------------------------------------------------------------***-------------------------------------------------------------------------- Abstract: Dermatology is one of the most unpredictable and difficult field to diagnose due its complexity. In the field of dermatology, many a times more tests need to be carried out so as to decide the skin condition the patient may be facing. The time may vary from practitioner to practitioner. This is also based on the experience of that person too. So, there is a need of a system which can detect the skin diseases without any of these constraints. Machine learning and image classifier can be used to efficiently detect the skin diseases. This paper is a study on the various techniques we can employ for the classification of skin diseases. This paper likewise briefs about the openly accessible skin dataset for testing and investigation. Keywords: Convolutional neural networks, Skin Disease, Tensorflow, machine learning I. INTRODUCTION Skin diseases are a major health problem in both high and low income countries and are the fourth leading cause of non-fatal skin disease burden. Skin diseases occur due to several factors like exposure to UV radiation, tanning, history of the family, environmental factors, alcohol etc. These factors affect the skin and have a devastating impact on its well-being. Skin diseases cause several problems like isolation, physical impairment, self-harm, body changes, difficulty in a relationship, unemployment, alcoholism and even death in case of malignant melanoma. Sometimes patients suffering from skin diseases attempts suicide. Classifying skin diseases require domain expertise, specialized equipment and expert knowledge and there is a gross mismatch between the burden of the skin patients and resources required to manage them. Especially people living in low- income countries do not have access to these resources. Therefore, to decrease the problems caused by skin diseases, there is a need for intelligent expert systems that can perform multi-class skin disease classification to help the people for early diagnosis. Image processing and machine learning based studies are being used in several areas such as face recognition, fingerprint recognition, tumor detection and segmentation. Different Machine Learning algorithms are used for the classification tasks in these areas. The commonly used ML algorithms are LDA, SVM, Artificial Neural Networks (ANN), Naive Bayes Classifier, K- Nearest Neighbor (KNN), and Deep learning algorithms. The selection of input feature is very important in any classification task, using ML algorithms. A. Skin Diseases Skin diseases vary in severity and symptoms. Some diseases can be permanent or temporary, and may be painful or painless. Some may be genetic, while other may have situational causes.. Some skin conditions are minor and others can be major. There are many different types of skin diseases, some are given below Acne These are commonly located on the face, neck, shoulders, chest, and upper back. Breakouts on the skin composed of blackheads, whiteheads, pimples, or deep, painful cysts and nodules and may leave scars or darken the skin if untreated. Psoriasis These are scaly, silvery, sharply defined skin patches and commonly located on the scalp, elbows, knees, and lower back. It may be itchy or asymptomatic.