We investigate Stochastic Mirror Descent (SMD) with matrix parameters and vector-valued predictions, a framework relevant to multi-class classification and matrix completion problems. Focusing on the overparameterized regime, where the total number of parameters exceeds the number of training samples, we prove that SMD with matrix mirror functions $ψ(\cdot)$ converges exponentially to a global interpolator. Furthermore, we generalize classical implicit bias results of vector SMD by demonstrating that the matrix SMD algorithm converges to the unique solution minimizing the Bregman divergence induced by $ψ(\cdot)$ from initialization subject to interpolating the data. These findings reveal how matrix mirror maps dictate inductive bias in high-dimensional, multi-output problems.

We investigate Gaussian Universality for data distributions generated via diffusion models. By Gaussian Universality we mean that the test error of a generalized linear model$f(\bf {W})$trained for a classification task on the diffusion data matches the test error of$f(\bf {W})$trained on the Gaussian Mixture with matching means and covariances per class. In other words, the test error depends only on the first and second order statistics of the diffusion-generated data in the linear setting. As a corollary, the analysis of the test error for linear classifiers can be reduced to Gaussian data from diffusion-generated data. Analyzing the performance of models trained on synthetic data is a pertinent problem due to the surge of methods such as [1]. Moreover, we show that, for any 1- Lipschitz scalar function$\phi$,$\phi (\mathbf {x})$is close to$\mathbb {E}\phi (\bf {x})$with high probability for$\bf {x}$sampled from the conditional diffusion model corresponding to each class. Finally, we note that current approaches for proving universality do not apply to diffusion-generated data as the covariance matrices of the data tend to have vanishing minimum singular values, contrary to the assumption made in the literature. This leaves extending previous mathematical universality results as an intriguing open question.

A celebrated result by Gordon allows one to compare the min-max behavior of two Gaussian processes if certain inequality conditions are met. The consequences of this result include the Gaussian min-max (GMT) and convex Gaussian min-max (CGMT) theorems which have had far-reaching implications in high-dimensional statistics, machine learning, non-smooth optimization, and signal processing. Both theorems rely on a pair of Gaussian processes, first identified by Slepian, that satisfy Gordon’s comparison inequalities. In this paper, we identify a new pair of Gaussian processes satisfying these inequalities. The resulting theorems extend the classical GMT and CGMT Theorems from the case where the underlying Gaussian matrix in the primary process has iid rows to where it has independent but non-identically-distributed ones. The new CGMT is applied to the problems of multi-source Gaussian regression, as well as to binary classification of general Gaussian mixture models.

This paper is eligible for the student paper award. Regularized linear regression is a promising approach for binary classification problems in which the training set has noisy labels since the regularization term can help to avoid interpolating the mislabeled data points. In this paper we provide a systematic study of the effects of the regularization strength on the performance of linear classifiers that are trained to solve binary classification problems by minimizing a regularized least-squares objective. We consider the over-parametrized regime and assume that the classes are generated from a Gaussian Mixture Model (GMM) where a fraction$c < \displaystyle \frac{1}{2}$of the training data is mislabeled. Under these assumptions, we rigorously analyze the classification errors resulting from the application of ridge,$\ell_{1}$, and$\ell_{\infty}$regression. In particular, we demonstrate that ridge regression invariably improves the classification error. We prove that$\ell_{1}$regularization induces sparsity and observe that in many cases one can sparsify the solution by up to two orders of magnitude without any considerable loss of performance, even though the GMM has no underlying sparsity structure. For$\ell_{\infty}$regularization we show that, for large enough regularization strength, the optimal weights concentrate around two values of opposite sign. We observe that in many cases the corresponding “compression” of each weight to a single bit leads to very little loss in performance. These latter observations can have significant practical ramifications. The full version of the manuscript containing full proofs and additional plots can be found in [1].

We study the problem of transfer learning and fine-tuning in linear models for both regression and binary classification. In particular, we consider the use of stochastic gradient descent (SGD) on a linear model initialized with pretrained weights and using a small training data set from the target distribution. In the asymptotic regime of large models, we provide an exact and rigorous analysis and relate the generalization errors (in regression) and classification errors (in binary classification) for the pretrained and fine-tuned models. In particular, we give conditions under which the fine-tuned model outperforms the pretrained one. An important aspect of our work is that all the results are "universal", in the sense that they depend only on the first and second order statistics of the target distribution. They thus extend well beyond the standard Gaussian assumptions commonly made in the literature. Furthermore, our universality results extend beyond standard SGD training to the test error of a classification task trained using ridge regression.