Multilinear subspace learning (MSL) aims to learn a specific small part of a large space of multidimensional objects having a particular desired property. It is a  dimensionality reduction approach for finding a low-dimensional representation with certain preferred characteristics of high-dimensional tensor data through direct mapping, without going through vectorization. The term tensor in MSL refers to multidimensional arrays. Examples of tensor data include images (2D/3D), video sequences (3D/4D), and hyperspectral cubes (3D/4D). The mapping from a high-dimensional tensor space to a low-dimensional tensor space or vector space is named as multilinear projection.