Hybridization of Decision Trees with other Techniques - Data Mining with Decision Trees: Theory and Applications - page 246

Fig. 15.2 The cross-inspection procedure outline: Searching for statistical significance,
the procedure compares the accuracy estimations of two distinct classifiers.

group. Finally, using three synthetic datasets, CPOM distinguished between
different populations in an underlined dataset.

15.4 Induction of Decision Trees by an Evolutionary

Algorithm (EA)

EAs are stochastic search algorithms inspired by the concept of Darwinian
evolution. The motivation for applying EAs to data mining tasks is that
they are robust, adaptive search techniques that perform a global search in
the solution space [ Freitas (2005) ] . Since a well-designed EA continually
considers new solutions, it can be viewed as an “anytime” learning
algorithm capable of quite quickly producing a good-enough solution. It
then continues to search the solution space, reporting the new “best”
solution whenever one is found.