A Review Paper on Classification of Genetic Algorithms

Abstract

Characteristic selection approaches were widely implemented to deal with the small pattern
length hassle within the analysis of micro-array datasets. For the multiclass problem, the
proposed strategies are based totally on the idea of selecting a gene subset to differentiate all
instructions. But, it is going to be extra powerful to solve a multiclass hassle through splitting
it into a set of -class problems and fixing each trouble with a respective classification
machine.In application, a multiclass hassle is split into a hard and fast of -magnificence
problems, every of that is tackled by means of a SE first. Powerful methods are proposed to
remedy the issues bobbing up within the fusion of SEs, and a greedy set of rules is designed
to preserve high diversity in SEs. This GP is tested in five datasets. The effects display that
the proposed method successfully implements the feature choice and class tasks.