Some enhancements in genetic learning: A case study on initial population

Abstract

In our project, we have selected the paradigm of genetic algorithms (GAs) for search through a space of all possible concept descriptions. We designed and implemented a system that integrates a domain-independent GA into the covering learning algorithm CN4, a large extension of the well-known CN2; we call it GA-CN4. This paper focuses on some enhancements of the above learning algorithm. Particularly, (a) discretization and fuzzification of numerical attributes for GAs (their genuine form is not able to process these attributes), (b) representation of chromosomes (objects, individuals) in GAs, and (c) the ways of initialization of a population for a GA.