Regression and Classification
y = f(x)
In case of supervised learning, depending on the type of output y , we can divide supervised learning broadly into two branches. If y belongs to a continuous range ( whether it be R ( set of real numbers ) or [a ,b ] or (a ,b ) or [a ,b ) or (a ,b ] where a ,b belong to R ), it comes under the branch of Regression. If y belongs to set of categorical values ( whether it be binary or some set of N ( natural numbers ), or some closed or semi-closed range of natural numbers, or some set of classes ), it comes under the branch of Classification.
This figure nicely depicts classification and regression. In both scenarios, we had some initial data points. In case of classification, we need a mapping to predict whether a data point belongs to category + or category o . One simple answer is points to the right (above ) of (the) line belong to + class and the points left (below) of (the) line belong to o class. In case of regression, we need to predict precise location of point given some input information. Here, our ouput location is in a continuos range, infact, this is a plot of linear regression.