Swarm-based optimization algorithms have demonstrated to have effective ability to solve the classification problem in multiclass databases. However, these algorithms tend to suffer from premature convergence in the high dimensional problem space.
This paper proposes a novel simplified swarm optimization (SSO) algorithm to overcome the above convergence problem by incorporating it with the new local search strategy. The proposed algorithm can find a better solution from the neighbourhood of the current solution produced by SSO. The performance of the proposed algorithm has been evaluated by using 13 different widely used databases and compared with the standard PSO and three other well-known classification algorithms. In addition, the practicability of the approach is studied by applying it in analysing golf swing from weight shift data. Empirical results illustrate that the proposed algorithm can achieve the highest classification accuracy.