Linear Regression - Essentials

• There must be a linear relationship between the independent variable and the dependent variable
• Note multicollinearity, autocorrelation, and heteroskedasticity among multiple variables .
• Linear regression is very sensitive to outliers. It can seriously affect the regression line and ultimately the predicted value.
• Multicollinearity increases the variance of the coefficient estimates, making the estimates very sensitive to small changes in the model. The result is that the coefficient estimates are not stable
• In the case of multiple independent variables, we can use forward selection, backward elimination, and stepwise screening to select the most important independent variable.