Tensor network algorithms for large scale learning and simulation

The novo nordisk data science collaborative research grant on quantum inspired is aimed at developing a
world class research program on tensor network algorithms and applications. The research program will be carried out with two ITU partners (myself and Nutan Limaye) and two external partners: Morten Mørup (DTU) and Raghavendra Selvan (KU). The aim is threefold: (i) to explore the use of tensor network constructions to dramatically reduce the memory
bottleneck of various machine learning tasks, with a special emphasis on life science applications, (ii) to develop tensor network simulation algorithms for
quantum circuit simulations, and (iii) to explore the fundamental limitations of tensor networks for learning tasks. The three aims will be tied together by a universal set of tools that might find further applications elsewhere.