The importance of data classification using machine learning methods in microarray data

Abstract

The detection of genetic mutations has attracted global attention. several methods have proposed to detect diseases such as cancers and tumours. One of them is microarrays, which is a type of representation for gene expression that is helpful in diagnosis. To unleash the full potential of microarrays, machine-learning algorithms and gene selection methods can be implemented to facilitate processing on microarrays and to overcome other potential challenges. One of these challenges involves high dimensional data that are redundant, irrelevant, and noisy. To alleviate this problem, this representation should be simplified. For example, the feature selection process can be implemented by reducing the number of features adopted in clustering and classification. A subset of genes can be selected from a pool of gene expression data recorded on DNA micro-arrays. This paper reviews existing classification techniques and gene selection methods. The effectiveness of emerging techniques, such as the swarm intelligence technique in feature selection and classification in microarrays, are reported as well. These emerging techniques can be used in detecting cancer. The swarm intelligence technique can be combined with other statistical methods for attaining better results.