Classification using cultural co-evolution and genetic programming
Myriam Abramson, Lawrence Hunter · 1996
Genetic programming has been used to evolve decision trees for classification [Koza, 1992] but its role as an inductive learner remains elusive. This paper examines a possible role of genetic programming as a learner in a cultural co-evolution model of learners developed by Hunter. We demonstrate that this cooperation allows classification to evolve much faster than their individual components and report encouraging results on a class of problems.