期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2013
卷号：4
期号：4
页码：62-73
语种：English
出版社：Ayushmaan Technologies
摘要：Cancer classification using gene expression data stands out from the other previous classification data due to its unique nature and application domain. Hope to gain some insight into the problem of cancer classification in aid of further developing more effective and efficient classification algorithms. With the development of Microarray technology, number of cancer can be identified. In Previous numbers of techniques are used, but they didn’t show the better accuracy.A novel approach to combine feature (gene) selection and Transductive Extreme Learning Machine (TELM) is proposed.The selected genes of the microarray data were then exploited to design the TELM. In this paper two types of future selection are taken for cancer classification they are (CBFS) Consistency-Based Feature Selection and Signal- Noise Ratio (SNR) with the help of Leukemia and Lymphoma Data Set.Experimental resultsconfirm the effectiveness of the proposed technique compared tothe TSVM (Transductive Extreme Learning Machine) cancer classification as well as gene-marker identification.
关键词：Cancer Classification;Transductive Extreme Learning Machine; Consistency-Based Feature Selection,Signal- Noise Ratio; Leukemia and Lymphoma Data Set.
