Tensor-Based Methods for Dimension Reduction of Volumetric Data

摘要

Tensor-based methods enjoy linear reduction of hyper-volumetric data. In this paper, we deal with tensor-to-tensor projection for volumetric data. The tensor-to-tensor projections transform a tensor to the products of lower-dimensional tensors. The dimension reduction based on the tensor-to-tensor projections is mathematically equivalent to dimension reduction in a vector space.