arXiv Statistics ML

Local polynomial density ratio estimation

arXiv:2609. 38412v1 Announce Type: cross Abstract: We propose a novel local-polynomial estimator of the ratio $r=f/g$ of two $d$-dimensional densities $f$ and $g$, from which independent samples are available.

arXiv Machine Learning
Oct 7

Sample-Optimal Estimation of the Fr\'echet Inception Distance

The paper investigates the sample complexity of estimating the Fréchet Inception Distance (FID) between high‑dimensional Gaussian distributions when one distribution is known. It establishes tight bias and variance bounds for the empirical plug‑in estimator, shows how higher‑order extrapolation reduces bias, and introduces a new Relative Taylor Debiasing (RTD) algorithm that achieves optimal sample complexity. Experiments on synthetic data and ImageNet confirm the theoretical predictions and demonstrate RTD’s superior accuracy with limited samples.

By Ziyun Chen, Jerry Li, Kevin Tian, Yusong Zhu
arXiv Machine Learning
Jul 28

Minimax Lower Bounds of Kernel Discrepancy Estimation: MMD, HSIC, KSD

arXiv:2607. 24235v1 Announce Type: cross Abstract: Over the past 20 years, kernel discrepancies have been leveraged as a highly powerful tool for quantifying the disagreement of distributions, with numerous successful applications in two-sample, goodness-of-fit, and independence testing, among others.

By Jose Cribeiro-Ramallo, Florian Kalinke, Zolt\'an Szab\'o
arXiv Machine Learning
Oct 16

Computational and Statistical Hardness of Calibration Distance

The paper investigates the computational and statistical challenges of the calibration distance, a key metric for measuring miscalibration in probabilistic predictors. It presents an efficient algorithm for exact computation when the distribution has a uniform marginal and noiseless labels, and shows that the problem becomes NP‑hard if either assumption is dropped. The authors also provide a polynomial‑time approximation scheme for the general case and establish tight sample complexity bounds for estimating the calibration distance from data.

By Mingda Qiao
arXiv Machine Learning
Sep 25

On the SoS Certifiability of Log-Concave Distributions

arXiv:2609. 30105v1 Announce Type: new Abstract: For an arbitrary isotropic log-concave distribution $P$ on $\mathbb{R}^d$, we prove that the polynomial $(Cm)^m\|v\|_2^m - \mathbb{E}_{X\sim P}\langle X,v\rangle^m$ is a sum of squares for every even $m\ge2$, where $C>0$ is a universal constant.

By Aleksandr Storozhenko