What is it about?
Symbolic regression is a family of algorithms that aims to find numerical predictors, for example, given various features (house size, house area, number of rooms) predict the house's price. We present a number of evolutionary algorithms that convert a given numerical predictor into a performant classifier (for example, instead of predicting a house's price, predict whether the price is high or low).
Featured Image
Photo by Anne Nygård on Unsplash
Why is it important?
Among others, we successfully employ a concept known as coevolution, where several populations coevolve simultaneously.
Perspectives
I hope the algorithms I've designed will be deemed a worthy addition to the Machine Learning toolkit of classification algorithms.
Professor Moshe Sipper
Ben-Gurion University of the Negev
Read the Original
This page is a summary of: Binary and multinomial classification through evolutionary symbolic regression, July 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3520304.3528922.
You can read the full text:
Resources
Contributors
The following have contributed to this page
