Multicategory Classification Using Relevance Vector Machine for Microarray Gene Expression Cancer Diagnosis

Main Article Content

Mrs. S. Sasikala
Dr. S. Santhosh Baboo


This paper deals with the advancement in cancer multicategory classification using Relevance Vector Machine (RVM) for
microarray gene expression cancer diagnosis. The proposed technique can be highly used for directing multicategory classification problems in
the cancer diagnosis area. SVM and ELM are the presently available techniques used for binary classification tasks, which is related to and
contains elements of non-parametric applied statistics, neural networks and machine learning. The cancer classification using the present
approach does not provide the expected accuracy and sometimes the result of clustering may be wrong. To overcome this problem an efficient
cancer classification using the Relevance Vector Machine (RVM) is proposed in this paper. This learning algorithm can generate accurate and
robust classification results on a sound theoretical basis, even when input data are non-monotone and non-linearly separable. The performance of
RVM is evaluated for the multicategory classification on two benchmark microarray data sets for cancer diagnosis, namely, the Lymphoma and
Leukemia dataset. The results indicate that RVM produces better classification accuracies than the approach using SVM and ELM when the data
given as input are preprocessed. RVM delivers very high performance with reduced training time and implementation complexity is less when
compared to artificial neural networks methods like conventional back-propagation ANN and Linder’s SANN.


Keyword- RVM, SVM, ELM, ANOVA, Cancer Classification and Gene Expression


Download data is not yet available.

Article Details