Affiliations: Department of Electrical Engineering, Faculty of Engineering, University of Birjand, Birjand, Iran
Correspondence:
[*]
Corresponding author: Hadi Shahraki, Department of Electrical Engineering, Faculty of Engineering, University of Birjand, Birjand, Iran. E-mail:hadi.shahraki@birjand.ac.ir
Abstract: In this paper a new classifier based on ant colony optimization in
continuous domains (ACOR) has been proposed. In this classifier,
ACOR is used to find the decision hyperplanes between the different
classes. Also, a new cost function is proposed to find maximum-margin
hyperplanes. The effectiveness of proposed classifier is evaluated by some
common benchmarks with different feature space dimensions and number of
classes. Also the performance of proposed classifier is compared with some
of the conventional, heuristic and ACO-based classifier. The experimental
results show that the performance of proposed classifier is comparable and
sometimes better than compared classifiers.
