Patterns are of central importance in DM and knowledge discovery. DM algorithms search the given data for patterns. Discovered patterns that are valid, interesting, and useful can be called knowledge.
Frawley etal. define a pattern in a data set as a statement that describes relationships in a subset of the data set with some certainty, such that the statement is simpler (in some sense) than the enumeration of all facts in the data set. A pattern thus splits the data set, as it pertains to a part of it, and may involve a spatial aspect which can be visualized.
This section introduces the most common types of patterns that are considered by DM algorithms. Note that the same type of pattern may be used in different DM algorithms addressing different tasks: trees can be used for classification, regression, or clustering (conceptual), and so can distance-based patterns.
Was this article helpful?