Low rank structure is prevalent in many different disciplines from data science and engineering, and taking advantage of it often allows for better visualization, for inferring hidden information, and for more efficient computations. An archetypical problem is to approximate a matrix with another one of low rank, which is known as principal component analysis. In practice, it is often the case that some entries of the matrix are unknown and have to be filled in (matrix completion). It is also common to impose additional constraints on the matrix. In other cases one needs to rely on generalized notions of low rank structure, such as tensors. This mini-symposium centers around low rank models in a broad sense, and covers computational, theoretical, and practical aspects.

All times are in Toronto local time (Eastern Time).