Keeping Track of Revised Parameters

Based on what you've learned during model assessment, it's time to have another look at the models. You have two options here:

• Adjust the parameters of existing models.
• Choose a different model to address your data mining problem.

In both cases, you'll be returning to the building models task and iterate until the results are successful. Don't worry about repeating this step. It is extremely common for data miners to evaluate and rerun models several times before finding one that meets their needs. This is a good argument for building several models at once and comparing the results before adjusting the parameters for each.