Robust low-rank tensor regression via truncation and adaptive Huber loss

05/03/2022

∙

by   Kangqiang Li, et al.

∙

0

∙

share

This paper investigates robust low-rank tensor regression with only finite (1+ϵ)-th moment noise based on the generalized tensor estimation framework proposed by Han et al. (2022). The theoretical result shows that when ϵ≥ 1, the robust estimator possesses the minimax optimal rate. While 1> ϵ>0, the rate is slower than the deviation bound of sub-Gaussian tails.