Erez Buchweitz | Searcharxiv

arXiv2025

Asymmetric Penalties Underlie Proper Loss Functions in Probabilistic Forecasting

Accurately forecasting the probability distribution of phenomena of interest is a classic and ever more widespread goal in statistics and decision theory. In comparison to point forecasts, probabilistic forecasts aim to provide a more complete and informative characterization of the target variable. This endeavor is only fruitful, however, if a forecast is "close" to the distribution it attempts to predict. The role of a loss function -- also known as a scoring rule -- is to make this precise by providing a quantitative measure of proximity between a forecast distribution and target random variable. Numerous loss functions have been proposed in the literature, with a strong focus on proper losses, that is, losses whose expectations are minimized when the forecast distribution is the same as the target. In this paper, we show that a broad class of proper loss functions penalize asymmetrically, in the sense that underestimating a given parameter of the target distribution can incur larger loss than overestimating it, or vice versa. Our theory covers many popular losses, such as the logarithmic, continuous ranked probability, quadratic, and spherical losses, as well as the energy and threshold-weighted generalizations of continuous ranked probability loss. To complement our theory, we present experiments with real epidemiological, meteorological, and retail forecast data sets. Further, as an implication of the loss asymmetries revealed by our work, we show that hedging is possible under a setting of distribution shift.

math.ST↗

arXiv2020

Two-Stage Regularization of Pseudo-Likelihood Estimators with Application to Time Series

Estimators derived from score functions that are not the likelihood are in wide use in practical and modern applications. Their regularization is often carried by pseudo-posterior estimation, equivalently by adding penalty to the score function. We argue that this approach is suboptimal, and propose a two-staged alternative involving estimation of a new score function which better approximates the true likelihood for the purpose of regularization. Our approach typically identifies with maximum a-posteriori estimation if the original score function is in fact the likelihood. We apply our theory to fitting ordinary least squares (OLS) under contemporaneous exogeneity, a setting appearing often in time series and in which OLS is the estimator of choice by practitioners.

stat.ME↗

arXiv2017

Concentration between Lévy's inequality and the Poincaré inequality for log-concave densities

Given a suitably normalized $X\in\mathbb{R}^n$ we observe that the function $θ\mapsto\mathbb{E}|X\cdotθ|$, defined for $θ\in S^{n-1}$, admits surprisingly strong concentration far surpassing what is expected on account of Lévy's isoperimetric inequality. Among the measures to which the above holds are all log-concave measures, for which a solution of the similar problem concerning the third marginal moments $θ\mapsto\mathbb{E} (X\cdot θ)^3$ would imply the hyperplane conjecture.

math.FA↗