A discrete particle swarm optimizer for graphic presentation of GMDH network

Particle swarm optimization is a new population based optimization methodology. With a characteristic of high performance and easy implementation, the particle swarm optimizer has many successful cases. However, it still has a problem of efficiency in solving sequential problems. In this paper, we propose a more efficient discrete particle swarm algorithm in solving the sequential problems such as graphic presentation of GMDH network. The dataset of Fisher´s Iris Plant is tested, and the results are shown.