no code implementations • 29 May 2023 • Yubo Zhuang, Xiaohui Chen, Yun Yang, Richard Y. Zhang
In contrast, nonnegative matrix factorization (NMF) is a simple clustering algorithm widely used by machine learning practitioners, but it lacks a solid statistical underpinning and theoretical guarantees.
no code implementations • 26 May 2023 • Gavin Zhang, Hong-Ming Chiu, Richard Y. Zhang
Recently, the technique of preconditioning was shown to be highly effective at accelerating the local convergence of non-convex gradient descent when the measurements are noiseless.
1 code implementation • 30 Nov 2022 • Hong-Ming Chiu, Richard Y. Zhang
Nevertheless, once a model has been adversarially trained, one often desires a certification that the model is truly robust against all future attacks.
no code implementations • 23 Oct 2022 • Geyu Liang, Gavin Zhang, Salar Fattahi, Richard Y. Zhang
This paper focuses on complete dictionary learning problem, where the goal is to reparametrize a set of given signals as linear combinations of atoms from a learned dictionary.
1 code implementation • 24 Aug 2022 • Gavin Zhang, Hong-Ming Chiu, Richard Y. Zhang
The matrix completion problem seeks to recover a $d\times d$ ground truth matrix of low rank $r\ll d$ from observations of its individual elements.
no code implementations • 5 Jul 2022 • Richard Y. Zhang
We consider minimizing a twice-differentiable, $L$-smooth, and $\mu$-strongly convex objective $\phi$ over an $n\times n$ positive semidefinite matrix $M\succeq0$, under the assumption that the minimizer $M^{\star}$ has low rank $r^{\star}\ll n$.
no code implementations • 7 Jun 2022 • Gavin Zhang, Salar Fattahi, Richard Y. Zhang
We consider using gradient descent to minimize the nonconvex function $f(X)=\phi(XX^{T})$ over an $n\times r$ factor matrix $X$, in which $\phi$ is an underlying smooth convex cost function defined over $n\times n$ matrices.
no code implementations • 21 Apr 2021 • Richard Y. Zhang
Under the restricted isometry property (RIP), we prove, for the general overparameterized regime with $r^{\star}\le r$, that an RIP constant of $\delta<1/(1+\sqrt{r^{\star}/r})$ is sufficient for the inexistence of spurious local minima, and that $\delta<1/(1+1/\sqrt{r-r^{\star}+1})$ is necessary due to existence of counterexamples.
no code implementations • NeurIPS 2020 • Gavin Zhang, Richard Y. Zhang
Optimizing the threshold over regions of the landscape, we see that for initial points around the ground truth, a linear improvement in the quality of the initial guess amounts to a constant factor improvement in the sample complexity.
no code implementations • NeurIPS 2020 • Richard Y. Zhang
If the relaxation is loose, however, then the resulting certificate can be too conservative to be practically useful.
1 code implementation • 19 Nov 2019 • Yi Ouyang, Richard Y. Zhang, Javad Lavaei, Pravin Varaiya
The offset optimization problem seeks to coordinate and synchronize the timing of traffic signals throughout a network in order to enhance traffic flow and reduce stops and delays.
Optimization and Control Systems and Control Systems and Control
no code implementations • 7 Jan 2019 • Richard Y. Zhang, Somayeh Sojoudi, Javad Lavaei
Using the technique, we prove that in the case of a rank-1 ground truth, an RIP constant of $\delta<1/2$ is both necessary and sufficient for exact recovery from any arbitrary initial point (such as a random point).
no code implementations • NeurIPS 2018 • Richard Y. Zhang, Cédric Josz, Somayeh Sojoudi, Javad Lavaei
When the linear measurements of an instance of low-rank matrix recovery satisfy a restricted isometry property (RIP)---i. e. they are approximately norm-preserving---the problem is known to contain no spurious local minima, so exact recovery is guaranteed.
no code implementations • ICML 2018 • Richard Y. Zhang, Salar Fattahi, Somayeh Sojoudi
The sparse inverse covariance estimation problem is commonly solved using an $\ell_{1}$-regularized Gaussian maximum likelihood estimator known as "graphical lasso", but its computational cost becomes prohibitive for large data sets.
no code implementations • 24 Nov 2017 • Salar Fattahi, Richard Y. Zhang, Somayeh Sojoudi
We have also derived a closed-form solution that is optimal when the thresholded sample covariance matrix has an acyclic structure.
1 code implementation • 10 Oct 2017 • Richard Y. Zhang, Javad Lavaei
Clique tree conversion solves large-scale semidefinite programs by splitting an $n\times n$ matrix variable into up to $n$ smaller matrix variables, each representing a principal submatrix of up to $\omega\times\omega$.
Optimization and Control