Togaware DATA MINING
Desktop Survival Guide
by Graham Williams
Google

C

C5.0: A commercial decision tree and rule induction product from www.rulequest.comRuleQuest developed by Ross Quinlan as the successor to his very successful and widely used ID3 and C4.5 systems.

Churn: The process of customer turnover, as occurs in many consumer service providers such as telecommunications and credit providers.

Classification:

Coevolution: An Evolutionary Computation paradigm where both the measure of fitness and the solution evolve separately and affect each other.

Computational Intelligence: Describes numerically based Artificial Intelligence systems distinguished from Symbolic Intelligence. Generally covers Evolutionary Computation, Neural Networks and Fuzzy Logic.

Cross Validation: A method of comparing or confirming performance of predictive models by estimating the error that would be produced by a model. Cross validation is also used to compare one method of inducing a predictive model (called an inducer) with another inducer. Cross validation can also be used to refine the parameters of a particular inducer. In this case the inducer's parameters are refined to minimise the error in the model produced. A k-fold cross validation will build k models from k datasets.

Confidence:

Cubist: A commercial piecewise-linear regression product from RuleQuest developed by Ross Quinlan. Uses decision tree induction to identify subsets of the data on which to perform regression.

Copyright © 2004-2006 Graham.Williams@togaware.com
Support further development through the purchase of the PDF version of the book.
Brought to you by Togaware.