Asymptotics of Ranks of Tensors

Matthias Christandl
University of Copenhagen

Unlike the matrix case, there are many different types of ranks for tensors. Unlike the matrix case, these types of ranks are rarely multiplicative, requiring us to look at amortized/regularized/asymptotic versions. Asymptotic ranks have important applications ranging from computational complexity to combinatorics to quantum information (my favorite corner of the world). I will present a bunch of results (and a conjecture!) on (1) weighted slice rank and (2) symmetric subrank that might amuse fans of Strassen, Comon or Tao.

Presentation (PDF File)

Back to Workshop IV: Efficient Tensor Representations for Learning and Computational Complexity