Extracting features is one of the most critical tasks in data mining, and it generally affects your end result more than the choice of data mining algorithm. Unfortunately, there are no hard and fast rules for choosing features that will result in high-performance data mining. The choice of features determines the model that you are using to represent your data.
Model creation is where the science of data mining becomes more of an art and why automated methods of performing data mining (there are several methods of this type) focus on algorithm choice and not model creation. Creating good models relies on intuition, domain expertise, data mining experience, trial and error, and sometimes a little luck.
