Tag
This paper introduces a low-rank degree-two density projection method for nonparametric changepoint detection in high dimensions, using matrix mean estimation to handle distributional changes without parametric assumptions.