Label encoding

In Machine Learning, most of the algorithms work better with numerical inputs. Therefore, one of the main challenge is to convert text/categorical data into numerical data and make an algorithm/model to make sense out of it. Moreover, neural networks expects input values to be numerical. That's why, we used LabelEncoder to normalize labels such that they contain only values between 0 and n_classes - 1.