A New Cluster-based Instance Selection Algorithm

The main contribution of the paper is proposing and evaluating, through the computational experiment, an agent-based population learning algorithm generating a representative training dataset of the required size. The proposed approach is based on the assumption that prototypes are selected from clusters. Thus, the number of clusters produced has a direct influence on the size of the reduced dataset. Agents within an A-Team execute various local search procedures and cooperate to find-out a solution to the instance reduction problem aiming at obtaining a compact representation of the dataset. Computational experiment has confirmed that the proposed algorithm is competitive to other approaches.