Data mining methods have many origins, including drawing on insights into learning as it naturally occurs in humans (cognitive science), and advances in computer science and algorithm design on how to best detect patterns in unstructured data. Although traditional statistical methods for analyzing data, based on statistical theories and models, are now widely accepted throughout various industries, data mining methods have only been widely embraced in business for a decade or two. However, their effectiveness for root cause analysis, and for modeling, optimizing and improving complex processes, are making data mining increasingly popular--and even necessary--in many real-world discrete manufacturing, batch manufacturing, and continuous-process applications.

There is no single, generally agreed-upon definition of data mining. As a practical matter, whenever data describing a process are available, in manufacturing for example, then any systematic review of those data to identify useful patterns, correlations, trends, and so forth, could be called “data mining.” Put simply, data mining uncovers nuggets of information from a sometimes vast repository of data describing the process of interest.

 …