Model evaluation is performed to ensure that a fitted model can accurately predict responses for future or unknown subjects. Without model evaluation, we might train models that over-fit in the training data. To prevent overfitting, we can employ packages, such as caret
, rminer
, and rocr
to evaluate the performance of the fitted model. Furthermore, model evaluation can help select the optimum model, which is more robust and can accurately predict responses for future subjects.
In the following chapter, we will discuss how one can implement a simple R script or use one of the packages (for example, caret
or rminer
) to evaluate the performance of a fitted model.