Applying Ant Colony Optimization to configuring stacking ensembles for data mining

作者：

Highlights：

• The approach applies Ant Colony Optimization in configuring stacking ensemble.

• It is compared with existing methods on 18 benchmark datasets.

• It is used in a real-world cost-sensitive data mining problem.

• It outperforms existing methods on many datasets and the data mining problem.

• The approach generates good cost-sensitive ensembles from ordinary learning algorithms.

摘要

•The approach applies Ant Colony Optimization in configuring stacking ensemble.•It is compared with existing methods on 18 benchmark datasets.•It is used in a real-world cost-sensitive data mining problem.•It outperforms existing methods on many datasets and the data mining problem.•The approach generates good cost-sensitive ensembles from ordinary learning algorithms.

论文关键词：ACO,Ensemble,Stacking,Metaheuristics,Data mining,Direct marketing

论文评审过程：Available online 20 November 2013.

论文官网地址：https://doi.org/10.1016/j.eswa.2013.10.063