Hotelling Deflation on Large Symmetric Spiked Tensors
This work addresses signal estimation in noisy tensor data, offering incremental theoretical insights into deflation mechanisms.
This paper analyzes the deflation algorithm's performance when estimating low-rank symmetric spikes in large tensors corrupted by Gaussian noise, providing a precise characterization of vector alignments and estimated weights under non-trivial spike component correlations.
This paper studies the deflation algorithm when applied to estimate a low-rank symmetric spike contained in a large tensor corrupted by additive Gaussian noise. Specifically, we provide a precise characterization of the large-dimensional performance of deflation in terms of the alignments of the vectors obtained by successive rank-1 approximation and of their estimated weights, assuming non-trivial (fixed) correlations among spike components. Our analysis allows an understanding of the deflation mechanism in the presence of noise and can be exploited for designing more efficient signal estimation methods.