A Hybrid Gene Selection and Classification Approach for Microarray Data Based on Clustering and PSO
In this paper, a novel hybrid approach based on clustering and particle swarm optimization (PSO) is proposed for gene selection and classification of microarray data. In this approach, PSO combining with clustering method are used to perform gene selection to reduce redundancy. Firstly, genes are partitioned into a certain number of clusters by using K-means, and then PSO is used to perform gene selection from the clustered genes. Because of its better generalization performance with much faster convergence rate than other learning algorithms for neural networks, extreme learning machine (ELM) is chosen to perform sample classification in the hybrid method. The proposed method selects less redundant interpretable genes as well as increases prediction accuracy. The efficiency and effectiveness of the proposed method is verified by extensive comparisons with other classic methods on some open microarray data.
KeywordsGene selection clustering particle swarm optimization extreme learning machine
Unable to display preview. Download preview PDF.
- 5.Kennedy, J., Eberhart, R.: Particle Swarm Optimization. In: IEEE International Conference on Neural Networks, vol. 4, pp. 1942–1948 (1995)Google Scholar
- 6.Shi, Y., Eberhary, R.C.: A Modified Particle Swarm Optimizer. Computational Intelligence 6, 69–73 (1998)Google Scholar
- 11.Mishra, D., Sahu, B.: Feature Selection for Cancer Classification: A Signal-to-noise Ratio Approach. International Journal of Scientific & Engineering Research 2(4) (2011)Google Scholar