I've recently come across a few encounters where people are using genetic programming or genetic algorithms to build "best" performing models.
gplearn is an example of genetic programming used for regression model building.
I understand that this is not something new, especially in the field of neural networks. However, with any method of feature selection, there should be a review of these methods. Are there any overall (statistical) reviews of how genetic algorithms should be used in model building?
I'd like to know how the statistical community views the use of these methods.
I've also reviewed posts like
for answers on this topic, but I am searching for a comprehensive review of the topic.