Multiple regression is a generalization of linear regression. Instead of a single dependent variable, there are n of them. Multiple regression seeks to find the coefficients to place an (n - 1)-dimensional hyperplane through the given data points (each observation being one dependent variable value, and n independent variable values).