Classifications provided another important branch for supervised learning techniques and generally they are used as classifiers to classify binary or multiclass objects based on predictors. The working principle of classification algorithms is very similar to those regression algorithms, but a key difference between them is that the regression is used to predict or estimate the desired responses represented as a continuous numeric values, and the classification algorithms are used to predict the desired response in terms of classes, either binary or multiclass objects.