Statistical Query Lower Bounds for Tensor PCA

08/10/2020
βˆ™
by   Rishabh Dudeja, et al.
βˆ™
13
βˆ™

In the Tensor PCA problem introduced by Richard and Montanari (2014), one is given a dataset consisting of n samples 𝐓_1:n of i.i.d. Gaussian tensors of order k with the promise that 𝔼𝐓_1 is a rank-1 tensor and 𝔼𝐓_1 = 1. The goal is to estimate 𝔼𝐓_1. This problem exhibits a large conjectured hard phase when k>2: When d ≲ n β‰ͺ d^k/2 it is information theoretically possible to estimate 𝔼𝐓_1, but no polynomial time estimator is known. We provide a sharp analysis of the optimal sample complexity in the Statistical Query (SQ) model and show that SQ algorithms with polynomial query complexity not only fail to solve Tensor PCA in the conjectured hard phase, but also have a strictly sub-optimal sample complexity compared to some polynomial time estimators such as the Richard-Montanari spectral estimator. Our analysis reveals that the optimal sample complexity in the SQ model depends on whether 𝔼𝐓_1 is symmetric or not. For symmetric, even order tensors, we also isolate a sample size regime in which it is possible to test if 𝔼𝐓_1 = 0 or 𝔼𝐓_1 β‰ 0 with polynomially many queries but not estimate 𝔼𝐓_1. Our proofs rely on the Fourier analytic approach of Feldman, Perkins and Vempala (2018) to prove sharp SQ lower bounds.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

βˆ™ 12/23/2016

Spectral algorithms for tensor completion

In the tensor completion problem, one seeks to estimate a low-rank tenso...
βˆ™ 10/31/2017

Effective Tensor Sketching via Sparsification

In this paper, we investigate effective sketching schemes via sparsifica...
βˆ™ 12/09/2014

Max vs Min: Tensor Decomposition and ICA with nearly Linear Sample Complexity

We present a simple, general technique for reducing the sample complexit...
βˆ™ 04/23/2020

Private Query Release Assisted by Public Data

We study the problem of differentially private query release assisted by...
βˆ™ 09/26/2019

Crowdsourcing via Pairwise Co-occurrences: Identifiability and Algorithms

The data deluge comes with high demands for data labeling. Crowdsourcing...
βˆ™ 11/11/2019

Nonconvex Low-Rank Symmetric Tensor Completion from Noisy Data

We study a noisy symmetric tensor completion problem of broad practical ...
βˆ™ 04/04/2018

Tight Query Complexity Lower Bounds for PCA via Finite Sample Deformed Wigner Law

We prove a query complexity lower bound for approximating the top r dime...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.