Disordered systems insights on computational hardness
In this review article we discuss connections between the physics of disordered systems,
phase transitions in inference problems, and computational hardness. We introduce two …
phase transitions in inference problems, and computational hardness. We introduce two …
A precise high-dimensional asymptotic theory for boosting and minimum--norm interpolated classifiers
A precise high-dimensional asymptotic theory for boosting and minimum-l1-norm
interpolated classifiers Page 1 The Annals of Statistics 2022, Vol. 50, No. 3, 1669–1695 …
interpolated classifiers Page 1 The Annals of Statistics 2022, Vol. 50, No. 3, 1669–1695 …
Computational barriers to estimation from low-degree polynomials
Computational barriers to estimation from low-degree polynomials Page 1 The Annals of
Statistics 2022, Vol. 50, No. 3, 1833–1858 https://doi.org/10.1214/22-AOS2179 © Institute of …
Statistics 2022, Vol. 50, No. 3, 1833–1858 https://doi.org/10.1214/22-AOS2179 © Institute of …
Lattice-based methods surpass sum-of-squares in clustering
Clustering is a fundamental primitive in unsupervised learning which gives rise to a rich
class of computationally-challenging inference tasks. In this work, we focus on the canonical …
class of computationally-challenging inference tasks. In this work, we focus on the canonical …
Sum-of-squares lower bounds for densest k-subgraph
Given a graph and an integer k, Densest k-Subgraph is the algorithmic task of finding the
subgraph on k vertices with the maximum number of edges. This is a fundamental problem …
subgraph on k vertices with the maximum number of edges. This is a fundamental problem …
Sum-of-squares lower bounds for sparse independent set
The Sum-of-Squares (SoS) hierarchy of semidefinite programs is a powerful algorithmic
paradigm which captures state-of-the-art algorithmic guarantees for a wide array of …
paradigm which captures state-of-the-art algorithmic guarantees for a wide array of …
Sub-exponential time Sum-of-Squares lower bounds for Principal Components Analysis
Abstract Principal Components Analysis (PCA) is a dimension-reduction technique widely
used in machine learning and statistics. However, due to the dependence of the principal …
used in machine learning and statistics. However, due to the dependence of the principal …
Non-gaussian component analysis via lattice basis reduction
Abstract Non-Gaussian Component Analysis (NGCA) is the following distribution learning
problem: Given iid samples from a distribution on $\R^ d $ that is non-gaussian in a hidden …
problem: Given iid samples from a distribution on $\R^ d $ that is non-gaussian in a hidden …
Semidefinite programs simulate approximate message passing robustly
M Ivkov, T Schramm - Proceedings of the 56th Annual ACM Symposium …, 2024 - dl.acm.org
Approximate message passing (AMP) is a family of iterative algorithms that generalize
matrix power iteration. AMP algorithms are known to optimally solve many average-case …
matrix power iteration. AMP algorithms are known to optimally solve many average-case …
Potential Hessian Ascent: The Sherrington-Kirkpatrick Model
We present the first iterative spectral algorithm to find near-optimal solutions for a random
quadratic objective over the discrete hypercube, resolving a conjecture of Subag [Sub21] …
quadratic objective over the discrete hypercube, resolving a conjecture of Subag [Sub21] …