Regression analysis (which is considered a supervised learning algorithm) dates back to the early 1800s when Gauss and Legendre utilized these techniques to measure the trajectories of the planets around the sun (https://en.wikipedia.org/wiki/Regression_analysis). The regression algorithm's usage is still going strong within the predictive analytics community due to its large base of literature and its ability to adapt to a wide range of problems.
Linear regression is the basic regression technique to use when your target variable is continuous. Linear regression is built upon the concept of ordinary least squares, and the functional form of the model is:
The preceding formula illustrates that linear regression models are additive, that is, the results of the prediction are calculated by summing the cross-product values of all of the independent variables, and then adding an intercept (first term of the formula), and error term (last term of the formula). That calculation...
