of the paper by Garvesh Raskutti, Ming Yuan and Han Chen

As the cost of generating data decreases, now we can get access to bigger data, both in size and in dimensions. We call arrays with more than 2 dimensions tensors. In this seminar, we will firstly introduce the notation and concept of tensors, their calculation and decompositions. After that, we will focus on multi-response tensor regression, and consider a general convex optimization approach under low-dimensional structural assumptions. Finally we will present a numerical study.