A gene selection approach for Diabetic retinopathy microarray data classification using Ant Colony Optimization
Journal of Ophthalmic and Optometric Sciences,
Vol. 3 No. 4 (2019),
5 October 2019,
Page 1-10
https://doi.org/10.22037/joos.v3i4.37115
This paper proposes an approach for gene selection in microarray data. The proposed approach consists of a primary filter approach using adjusted p-value apply to the result of the differently expressed genes analyses which reduces the initial genes and hence the search space and time complexity. Then, a heuristic approach which is based on ant colony optimization method (ACO) is used to find the set of genes which improve the classification accuracy. The selected genes from the last phase are evaluated using (receiver operating characteristic) ROC curve and the most effective while smallest feature subset is determined. The classifier which are evaluated in the proposed framework is K-nearest neighbor. The proposed approach is evaluated on a diabetic retinopathy microarray dataset. The experiment shows that with 9 highly informative genes, the proposed approach has achieved a high precision rate. In addition, the six selected genes have been found meaningfully in the biology texts in comparison with the state-of-the-art. The evaluations confirm that the proposed approach can find the smallest subset of genes while approaching the maximum accuracy.