N&O Seminar Nick Vannieuwenhoven

Nick Vannieuwenhoven from KU Leuven gives a seminar talk titled "Hadamard-Hitchcock decompositions".

19 Feb 2024 from 11 a.m. to 19 Feb 2024 noon CET (GMT+0100)
L016 and online


Nick Vannieuwenhoven

KU Leuven


Hadamard-Hitchcock decompositions

A Hadamard-Hitchcock decomposition of a multidimensional array is a decomposition that expresses the latter as an elementwise or Hadamard product of several tensor rank decompositions. Such decompositions can encode probability distributions that arise from statistical graphical models associated to complete bipartite graphs with one layer of observed random variables and one layer of hidden ones, usually called restricted Boltzmann machines. Nick establishes generic identifiability of Hadamard-Hitchcock decompositions by exploiting the reshaped Kruskal criterion for tensor rank decompositions. A flexible algorithm leveraging existing decomposition algorithms for tensor rank decomposition is introduced for computing a Hadamard-Hitchcock decomposition. Numerical experiments illustrate its computational performance and numerical accuracy. This is joint work with Alessandro Oneto.

Zoom-Link for online participation: https://cwi-nl.zoom.us/j/87542191297?pwd=QTIxYlNQUmQ5UTFKb0NSaVVsbGFJZz09