[1B5-GS-2] Machine learning

[1B5-GS-2-04] Multi-Body Modeling of Non-Negative Tensors

〇Kazu Ghalamkari^1,2, Mahito Sugiyama^1,2 (1. Graduate University for Advanced Studies, SOKENDAI, 2. National Institute of Informatics (NII))

Keywords:Tensor decomposition, Low-rank approximation , Tensor networks

Global optimization of tensor low-rank decomposition by reducing ranks is fundamentally difficult due to the non-convexity of the associated cost function. Instead of ranks, we use interactions between modes as an alternative concept of ranks and formulate convex tensor decomposition that controls mode interactions by minimizing the Kullback--Leibler divergence from input.