Multiple Linear Regression

This is when 2 or more predictor (independent) variables are used to estimate a response (dependant) variable. Many-to-one relationship.

In a MLR model, the coefficient shows the impact on the dependant variable (\(y\)) if that independent variable (\(x1\)) is increased by one. Assuming all other variables stay constant

Multicollinearity

• Happens when the independent variables depend on each other or influence each other.

• This makes it difficult to understand what is causing the results in the dependant variable.

• When adding an independent variable consider the relationship between that variable and other independent variables

• The best MLR model includes independent variables that are not correlated with each other, only with the dependant variable.

Overfitting

• Too many independent variables that expalin teh variance but aren’t useful to the model