Anzeige
Erschienen in:
2014 | OriginalPaper | Buchkapitel
A Novel Approach to Gene Selection of Leukemia Dataset Using Different Clustering Methods
verfasst von : P. Prasath, K. Perumal, K. Thangavel, R. Manavalan
Erschienen in: Computational Intelligence, Cyber Security and Computational Models
Verlag: Springer India
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
Gene datasets from microarray comprise large number of genes. Clustering is a widely used approach for grouping similar kind of genes. The main objective of this paper is to identify the optimal subset of genes from the leukemia dataset in order to classify the leukemia cancer. Different clustering approaches such as K-means (KM) clustering, fuzzy C-means (FCM) clustering, and modified K-means (MKM) clustering have been adopted in this research. The clusters obtained from these methods are further clustered using K-means sample-wise (by omitting class values), and the results are compared with ground truth value to evaluate the performance of the different clustering methods. The highly correlated genes are selected from the cluster that produces more accurate classification results. It is observed that the FCM (gene-wise clustering) with K-means (sample-wise clustering) produces better accuracy, and the resultant genes have been identified.