Article,
A Novel Approach to Design Classifier Using Genetic Programming
IEEE Transactions on Evolutionary Computation, 8 (2): 183--196 (April 2004)
DOI: doi:10.1109/TEVC.2004.825567
Abstract
We propose a new approach for designing classifiers
for a c-class c>=2 problem using Genetic Programming
(GP). The proposed approach takes an integrated view of
all classes when the GP evolves. A multi-tree
representation of chromosomes is used. In this context,
we propose a modified crossover operation and a new
mutation operation that reduces the destructive nature
of conventional genetic operations. A new concept of
unfitness of a tree is used to select trees for genetic
operations. This gives more opportunity for unfit trees
to become fit. A new concept of OR-ing chromosomes in
the terminal population is introduced, which enables us
to get a classifier with better performance. Finally, a
weight based scheme and a heuristic rule based scheme
characterising typical mistakes are used for conflict
resolution. The classifier is capable of saying ``don't
know'' when faced with unfamiliar examples. The
effectiveness of our scheme is demonstrated on several
real data sets.
