Online robust principal component analysis (RPCA) algorithms recursively decompose incoming data into low-rank and sparse components. However, they operate on vectors cannot directly be applied to higher-order arrays (e.g. video frames). In this paper, we propose a new online PCA algorithm that preserves the multi-dimensional structure of data. Our is based recently proposed tensor singular val...

In the probabilistic framework of multivariate regular variation, first order behavior heavy-tailed random vectors above large radial thresholds is ruled by a homogeneous limit measure. For high dimensional vector, reasonable assumption that support this measure concentrated on lower subspace, meaning certain linear combinations components are much likelier to be than others. Identifying subspa...