Data Mining Operations

• Affinity grouping - identifying those products purchased alongside each other
• Sequence discovery - identifying the patterns of some customers purchasing the same products regularly
• Classification - separation of a particular case into several classes e.g. socio-economic groups
• Estimation - discrete outcomes e.g. is it worth marketing one particular product to customers who have no history of buying that product
• Regression - predicting unknown values from a different set of attributes
• Clustering - assigning cases into one of several categories

These facts can be utilised to identify patterns that can be expressed as predicates. This is knowledge about customers.

Data Mining Methods

Partially replace the human analyst by automatically generating models for testing against a database. Models are refined if they do not fit and re-tested. Decision procedures that are automated mechanical methods.

For complex queries the human analyst would need many queries to obtain an answer. This is time consuming and dependent on the analysts skill. Relationships within data are complex. As a result hypothesis verification by SQL is an unsatisfactory method for discovering patterns in data.

Decision Trees

Basically a hierarchical structure where questions are posed at each node. Which branch is followed is dependent on the answer. A flow chart with questions! The decision trees represent the rules discussed in facts and rules.

Comments, suggestions, ideas to
Stuart Banner