Classification of skin disease using ensemble-based classifier

Cancerous skin disease such as melanoma and nevi typically results from environmental factors (such as exposure to sunlight) among other causes. The necessary tools needed for early detection of these diseases are still not a reality in most communities. In this paper, the framework is proposed to deal with the detection and classification of various skin diseases. The two techiques commonly used for reduction of dimensionality are feature extraction and feature selection. In feature extraction, features are extracted from original data using principal component analysis and linear discrimant analysis and then extracted feature is reduced by feature selection technique called Fisher ratio method in which the subset of sufficient features is selected for classification. This technique improves the performance and enhances the speed of classifier. The ensemble-based classifier such as Bayesian, self-organised map and support vector machine are used to classify the various skin diseases from the data set. The proposed technique achieves better accuracy and less execution time than existing approach.