Protein classification is a task of paramount importance in various fields of biology. Despite the great momentum of modern implementation of protein classification, machine learning techniques such as Random Forest and Neural Network could not always be used for several reasons: Data collection, unbalanced classification or labelling of the data.