A Perspective View of Mining Rules with Ant Colony Optimization Technique:Ant-miner

In these days of computational technology, rule based classifiers in data mining are gaining popularity and provides a good platform for researchers to use different metaheuristics techniques to solve non-deterministic problems in real-world scenario. This study provides the implementation of ant based algorithms in the classification task of data mining. We have done an experimental study by taking ten datasets and calculated predictive accuracy measures of Ant-miner, cAnt-miner, JRIP and PART as well as non-statistical measurement tests to validate the results. This study shows that Ant based algorithms outperforms over JRIP as well as PART algorithms.