Introduction
The objective of data mining is to understand and predict behavior. A retailer wants to know why people buy, and how to sell more. An educator wants to know what factors influence educational and professional success and how to help students learn and prepare for a career. A criminologist wants to understand the factors that lead to crime, and how to prevent crime.
Data miners often speak of valuable patterns in data, and powerful models. What makes a pattern valuable? It's valuable if it adds to our understanding of behavior. What makes a model useful? It's useful if it is effective at predicting behavior.
Data miners aim to identify influential factors that drive behavior. When we identify those driving factors, through exploration of patterns in the data, we can understand behavior. If we can describe, with a quantitative model, the relationship between driving factors and behavior, then we can predict behavior.
It's easy for a data miner to make a model; that takes little more...