Stochastic gradient descent in high dimensions for multi-spiked tensor PCA

TL;DR

This paper analyzes the dynamics of stochastic gradient descent in high-dimensional multi-spiked tensor PCA, identifying sample complexity and SNR conditions for efficient spike recovery, revealing a sequential elimination process.

Contribution

It extends the understanding of tensor PCA by characterizing the sample and SNR thresholds for multi-spike recovery using SGD, and introduces the concept of sequential elimination in the recovery process.

Findings

Full recovery requires sample size scaling as N^{p-2}.

Spikes are recovered sequentially through a threshold-based process.

Separated SNRs enable exact spike recovery, equal SNRs recover the subspace.

Abstract

We study the high-dimensional dynamics of online stochastic gradient descent (SGD) for the multi-spiked tensor model. This multi-index model arises from the tensor principal component analysis (PCA) problem with multiple spikes, where the goal is to estimate $r$ unknown signal vectors within the $N$-dimensional unit sphere through maximum likelihood estimation from noisy observations of a $p$-tensor. We determine the number of samples and the conditions on the signal-to-noise ratios (SNRs) required to efficiently recover the unknown spikes from natural random initializations. We show that full recovery of all spikes is possible provided a number of sample scaling as $N^{p-2}$, matching the algorithmic threshold identified in the rank-one case [Ben Arous, Gheissari, Jagannath 2020, 2021]. Our results are obtained through a detailed analysis of a low-dimensional system that describes the evolution of the correlations between the estimators and the spikes, while controlling the noise in the dynamics. We find that the spikes are recovered sequentially in a process we term "sequential elimination": once a correlation exceeds a critical threshold, all correlations sharing a row or column index become sufficiently small, allowing the next correlation to grow and become macroscopic. The order in which correlations become macroscopic depends on their initial values and the corresponding SNRs, leading to either exact recovery or recovery of a permutation of the spikes. In the matrix case, when $p=2$, if the SNRs are sufficiently separated, we achieve exact recovery of the spikes, whereas equal SNRs lead to recovery of the subspace spanned by them.