Improved estimation in tensor regression with multiple change-points

In this paper, we consider an estimation problem about the tensor coefficient in a regression model with multiple and unknown change-points. We generalize some recent findings five ways. First, studied is more general than one context of matrix parameter Second, develop asymptotic results estimators Third, construct class shrinkage that encompasses unrestricted estimator (UE) restricted (RE). Fourth, identities which are crucial deriving distributional risk (ADR) estimators. Fifth, show proposed perform better UE. The additional novelty established consists fact dependence structure errors as weak L2-mixingale. Finally, theoretical corroborated by simulation our methods applied to analyse MRI fMRI datasets.