 In this video, I'll be talking about a parametric regression method called linear regression. It means that the learning method uses a fixed equation form that establishes the relationship between the dependent output variable and the independent covariates. Parametric regression models take the general form y is equal to ffx plus epsilon, where y is the predicted output variable, ffx is an unknown function, and epsilon is the error term that is independent of the covariates x. Different regression models use different forms of the function f. In linear regression with one covariate x, it takes the form beta0 plus beta1x. The parametric form has thus reduced the problem of finding a relationship between the covariates x and the response variable y to determining two coefficients, beta0 and beta1.