An ant colony based system for data mining: applications to medical data

Abstract

This work describes an algorithm for rule discovery in databases called AntMiner. The objective of the algorithm is the extraction of classification rules to be applied to unseen data as a decision aid. The algorithm used to discover such rules is inspired in the behavior of a real ant colony, as well as some concepts of information theory and data mining. AntMiner was applied to medical databases to obtain classification rules. 1