The model we're looking to create will consist of the following form:
In this formula, the predictor variables (features) can be from 1 to n.
One of the critical elements that we'll cover here is the vital task of feature selection. Later chapters will include more advanced techniques.
Forward selection starts with a model that has zero features; it then iteratively adds features one at a time until achieving the best fit based on say the reduction in residual sum of squares or overall model AIC. This iteration continues until a stopping rule is satisfied for example, setting maximum p-values for ...