Let $(x_{i}, y_{i})_{i=1,\dots,n}$ denote independent samples from a general mixture distribution $\sum_{c\in\mathcal{C}}\rho_{c}P_{c}^{x}$, and consider the hypothesis class of generalized linear models $\hat{y} = F(\Theta^{\top}x)$. In this work, we investigate the asymptotic joint statistics of the family of generalized linear estimators $(\Theta_{1}, \dots, \Theta_{M})$ obtained either from (a) minimizing an empirical risk $\hat{R}_{n}(\Theta;X,y)$ or (b) sampling from the associated Gibbs measure $\exp(-\beta n \hat{R}_{n}(\Theta;X,y))$. Our main contribution is to characterize under which conditions the asymptotic joint statistics of this family depends (on a weak sense) only on the means and covariances of the class conditional features distribution $P_{c}^{x}$. In particular, this allow us to prove the universality of different quantities of interest, such as the training and generalization errors, redeeming a recent line of work in high-dimensional statistics working under the Gaussian mixture hypothesis. Finally, we discuss the applications of our results to different machine learning tasks of interest, such as ensembling and uncertainty

翻译：让我们( x), y ⁇ i} ⁇ i= 1,\\\ dots, n} 美元表示来自一般混合物分配的独立样本 $\ sum\\ c\ c\ in\ mathcal{C\\\ r ⁇ c} P ⁇ c ⁇ x} 美元, 并且考虑通用线性模型的假设等级 $\ hat{y} = F(\\\\ theta ⁇ top}) 美元 。 在这项工作中, 我们调查通用线性估测器家庭(\\ theta},\ dots,\ dots,\theta}n} 美元, 美元表示独立的样本样本 。 我们的主要贡献在于根据什么条件, 这个家族的匹配性联合统计 $( 在一种微弱的意义上) 来减少一个经验性风险 $\ hat{r\\\\\\ {r} (theta; X,y) 或 (b) 从相关的 Gibbs 量 中取样 $\\\\\\\\\\\\\\\\\\\\ lagal real real real ladeal real real real real real relividudestrational ridedudeal ridudestration) ex rideal ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex exx ex ex ex ex ex ex ex ex ex exx ex ex ex ex exx exx lax ex ex exfus exx exlix exx exx ex ex ex ex ex ex ex ex exfolx exfol exual a exx exx ex ex ex ex ex exx ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex