Many studies demonstrated that the DNA methylation, which occurs in the context of a CpG, has a strong correlation with diseases, including cancer. There is a strong interest in analyzing the DNA methylation data to find how to distinguish different subtypes of the tumor [1]. However, the conventional statistical methods are not suitable for analyzing the highly dimensional DNA methylation data with bounded support. DNA methylation is one of the most extensively studied epigenetic marks and is known to be implicated in a wide range of biological processes, including chromosome instability, Xchromosome inactivation, cell differentiation, cancer progression and gene regulation [4]. Identification of cancer subtypes plays an important role in revealing useful insights into disease pathogenesis and advancing personalized therapy. In order to explicitly capture the properties of the data, a deep neural network is used, which composes of several stacked binary restricted Boltzmann machines, to learn the low-dimensional deep features of the DNA methylation data.