competitive Three Echelon Supply Chain Network Design Considering Service Level and Disruption

Feature extraction is one of the most important parts in pattern recognition especially in classification. Developing new feature extraction methods can lead toward better classification with higher accuracy. Linear discriminant analysis is one of the most well-known methods in feature extraction. Up to now, various methods have been presented for optimizing linear discriminant analysis. Optimization of the objective function of linear discriminant analysis is of a great issue. By optimizing the objective function of linear discriminant analysis, the data points belonging to different classes are well-separated from each other. It means, the classification can be done with a higher accuracy. This paper aims to use a refined multigrouped particle swarm optimization algorithm for optimizing linear discriminant analysis. Experimental studies on several datasets show that the proposed method can yield better performance than multigrouped particle swarm optimization.