Data integration model for cancer subtype identification using Kernel Dimensionality Reduction-Support Vector Machine (KDR-SVM)

In this paper, an integration model of cancer patients data types such as microarray DNA and clinical data will be experimentally explored. The data of integration will be used for cancer subtype identification using kernel based classification methods which is the extension of Support Vector Machine (SVM) approach with Kernel Dimensionality Reduction (KDR). KDR-SVM method will be implemented in Lymphoma cancer database and the relevant clinical information. Data type representation will be modeled in an appropriate kernel matrix. The results of the experiment show that the KDR-10 dimensions and data integration can improve the accuracy of the identification of subtype cancer.