This chapter looks at improving a model, which presupposes that at least a model exists. Chapter 11 covered the process of mining the initial model. However, just because a model exists doesn't mean that it has no problems, or that it is the best or most appropriate model that the data permits. This chapter looks at the process of turning the initially created model into one that better fits the data and the business problem. Although most of the processes apply as much to improving an explanatory model as to improving a classificatory model, most of the issues here are addressed as if the model to be improved is classificatory except, of course, where the issues discussed are specific to one type of model or the other.

The process of improving a model can, for discussion purposes, be broken into two broad categories: discovering where the model has problems, and fixing the discovered problems.

This chapter is divided into two major, but very connected, sections that align with these two categories and their associated activities:

• The first major section discusses how to determine model performance by performing a number of checks. (The methodology section of Part IV notes these as discoveries.)

• The second major section covers what to do about the results of the checks, or, essentially, how to fix the model.

These two activities diagnosing model problems and applying remedies are both applied in order to refine the initial model. This chapter presents the second step in the mining/refining...