Equivalence of Approximate Message Passing and Low-Degree Polynomials in Rank-One Matrix Estimation

We consider the problem of estimating an unknown parameter vector $\mathit{\theta }\in {\mathbb{R}}^n$, given noisy observations $\mathit{Y}=\mathit{\theta }{\mathit{\theta }}^{\mathrm{\top }}\mathrm{/}\sqrt{n}+\mathit{Z}$ of the rank-one matrix $\mathit{\theta }{\mathit{\theta }}^{\mathrm{\top }}$, where $\mathit{Z}$ has independent Gaussian entries. When information is available about the distribution of the entries of $\mathit{\theta }$, spectral methods are known to be strictly sub-optimal. Past work characterized the asymptotics of the accuracy achieved by the optimal estimator. However, no polynomial-time estimator is known that achieves this accuracy. It has been conjectured that this statistical-computation gap is fundamental, and moreover that the optimal accuracy achievable by polynomial-time estimators coincides with the accuracy achieved by certain approximate message passing (AMP) algorithms. We provide evidence towards this conjecture by proving that no estimator in the (broader) class of constant-degree polynomials can surpass AMP.