We revisit the mathematical foundations of proper scoring rules (PSRs) and Bregman divergences and present their characteristic properties in a unified theoretical framework. In many situations it is preferable not to generate a PSR directly from its convex entropy on the unit simplex but instead by the sublinear extension of the entropy to the positive orthant. This gives the scoring rule simply as a subgradient of the extended entropy, allowing for a more elegant theory. The other convex extensions of the entropy generate affine extensions of the scoring rule and induce the class of functional Bregman divergences. We discuss the geometric nature of the relationship between PSRs and Bregman divergences and extend and unify existing partial results. We also approach the topic of differentiability of entropy functions. Not all entropies of interest possess functional derivatives, but they do all have directional derivatives in almost every direction. Relying on the notion of quasi-interior of a convex set to quantify the latter property, we formalise under what conditions a PSR may be considered to be uniquely determined from its entropy.
One of the most common methods for statistical inference is the maximum likelihood estimator (MLE). The MLE needs to compute the normalization constant in statistical models, and it is often intractable. Using unnormalized statistical models and replacing the likelihood with the other scoring rule are a good way to circumvent such high computation cost, where the scoring rule measures the goodness of fit of the model to observed samples. The scoring rule is closely related to the Bregman divergence, which is a discrepancy measure between two probability distributions. In this paper, the purpos…
Minimization of suitable statistical distances~(between the data and model densities) has proved to be a very useful technique in the field of robust inference. Apart from the class of $ϕ$-divergences of \cite{a} and \cite{b}, the Bregman divergence (\cite{c}) has been extensively used for this purpose. However, since the data density must have a linear presence in the cross product term of the Bregman divergence involving both the data and model densities, several useful divergences cannot be captured by the usual Bregman form. In this respect, we provide an extension of the ordinary Bregman …
In this paper, we introduce new classes of divergences by extending the definitions of the Bregman divergence and the skew Jensen divergence. These new divergence classes (g-Bregman divergence and skew g-Jensen divergence) satisfy some properties similar to the Bregman or skew Jensen divergence. We show these g-divergences include divergences which belong to a class of f-divergence (the Hellinger distance, the chi-square divergence and the alpha-divergence in addition to the Kullback-Leibler divergence). Moreover, we derive an inequality between the g-Bregman divergence and the skew g-Jensen d…
In this paper a new family of minimum divergence estimators based on the Bregman divergence is proposed. The popular density power divergence (DPD) class of estimators is a sub-class of Bregman divergences. We propose and study a new sub-class of Bregman divergences called the exponentially weighted divergence (EWD). Like the minimum DPD estimator, the minimum EWD estimator is recognised as an M-estimator. This characterisation is useful while discussing the asymptotic behaviour as well as the robustness properties of this class of estimators. Performances of the two classes are compared -- bo…
In previous work the authors defined the k-th order simplicial distance between probability distributions which arises naturally from a measure of dispersion based on the squared volume of random simplices of dimension k. This theory is embedded in the wider theory of divergences and distances between distributions which includes Kullback-Leibler, Jensen-Shannon, Jeffreys-Bregman divergence and Bhattacharyya distance. A general construction is given based on defining a directional derivative of a function $ϕ$ from one distribution to the other whose concavity or strict concavity influences the…
Prediction error is critical to assessing the performance of statistical methods and selecting statistical models. We propose the cross-validation and approximated cross-validation methods for estimating prediction error under a broad q-class of Bregman divergence for error measures which embeds nearly all of the commonly used loss functions in regression, classification procedures and machine learning literature. The approximated cross-validation formulas are analytically derived, which facilitate fast estimation of prediction error under the Bregman divergence. We then study a data-driven op…
Minimum divergence procedures based on the density power divergence and the logarithmic density power divergence have been extremely popular and successful in generating inference procedures which combine a high degree of model efficiency with strong outlier stability. Such procedures are always preferable in practical situations over procedures which achieve their robustness at a major cost of efficiency or are highly efficient but have poor robustness properties. The density power divergence (DPD) family of Basu et al.(1998) and the logarithmic density power divergence (LDPD) family of Jones…
Logarithmic score and information divergence appear in both information theory, statistics, statistical mechanics, and portfolio theory. We demonstrate that all these topics involve some kind of optimization that leads directly to the use of Bregman divergences. If a sufficiency condition is also fulfilled the Bregman divergence must be proportional to information divergence. The sufficiency condition has quite different consequences in the different areas of application, and often it is not fulfilled. Therefore the sufficiency condition can be used to explain when results from one area can be…
Divergence functions play a key role as to measure the discrepancy between two points in the field of machine learning, statistics and signal processing. Well-known divergences are the Bregman divergences, the Jensen divergences and the f-divergences. In this paper, we show that the symmetric Bregman divergence can be decomposed into the sum of two types of Jensen divergences and the Bregman divergence. Furthermore, applying this result, we show another sum decomposition of divergence is possible which includes f-divergences explicitly.
The paper introduces scaled Bregman distances of probability distributions which admit non-uniform contributions of observed events. They are introduced in a general form covering not only the distances of discrete and continuous stochastic observations, but also the distances of random processes and signals. It is shown that the scaled Bregman distances extend not only the classical ones studied in the previous literature, but also the information divergence and the related wider class of convex divergences of probability measures. An information processing theorem is established too, but onl…
Polygenic risk scores (PRS) have recently received much attention for genetics risk prediction. While successful for the Caucasian population, the PRS based on the minority population suffer from small sample sizes, high dimensionality and low signal-to-noise ratios, exacerbating already severe health disparities. Due to population heterogeneity, direct trans-ethnic prediction by utilizing the Caucasian model for the minority population also has limited performance. In addition, due to data privacy, the individual genotype data is not accessible for either the Caucasian population or the minor…
Using a trimming approach, we investigate a k-means type method based on Bregman divergences for clustering data possibly corrupted with clutter noise. The main interest of Bregman divergences is that the standard Lloyd algorithm adapts to these distortion measures, and they are well-suited for clustering data sampled according to mixture models from exponential families. We prove that there exists an optimal codebook, and that an empirically optimal codebook converges a.s. to an optimal codebook in the distortion sense. Moreover, we obtain the sub-Gaussian rate of convergence for k-means 1 $\…
A general class of Bayesian lower bounds when the underlying loss function is a Bregman divergence is demonstrated. This class can be considered as an extension of the Weinstein--Weiss family of bounds for the mean squared error and relies on finding a variational characterization of Bayesian risk. The approach allows for the derivation of a version of the Cramér--Rao bound that is specific to a given Bregman divergence. The new generalization of the Cramér--Rao bound reduces to the classical one when the loss function is taken to be the Euclidean norm. The effectiveness of the new bound is ev…
New Jersey Institute of Technology（新泽西理工学院）；Bremen University of Applied Sciences（不来梅应用技术大学）；Princeton University（普林斯顿大学）；Technion - Israel Institute of Technology（以色列理工学院）
A common way of characterizing minimax estimators in point estimation is by moving the problem into the Bayesian estimation domain and finding a least favorable prior distribution. The Bayesian estimator induced by a least favorable prior, under mild conditions, is then known to be minimax. However, finding least favorable distributions can be challenging due to inherent optimization over the space of probability distributions, which is infinite-dimensional. This paper develops a dimensionality reduction method that allows us to move the optimization to a finite-dimensional setting with an exp…
We study the logarithmic $L^{(α)}$-divergence which extrapolates the Bregman divergence and corresponds to solutions to novel optimal transport problems. We show that this logarithmic divergence is equivalent to a conformal transformation of the Bregman divergence, and, via an explicit affine immersion, is equivalent to Kurose's geometric divergence. In particular, the $L^{(α)}$-divergence is a canonical divergence of a statistical manifold with constant sectional curvature $-α$. For such a manifold, we give a geometric interpretation of its sectional curvature in terms of how the divergence b…
Preserving the robustness of the procedure has, at the present time, become almost a default requirement for statistical data analysis. Since efficiency at the model and robustness under misspecification of the model are often in conflict, it is important to choose such inference procedures which provide the best compromise between these two concepts. Some minimum Bregman divergence estimators and related tests of hypothesis seem to be able to do well in this respect, with the procedures based on the density power divergence providing the existing standard. In this paper we propose a new famil…
School of Liberal Arts, Korea University of Technology and Education（韩国技术教育大学文理学院）
Exponential dispersion model is a useful framework in machine learning and statistics. Primarily, thanks to the additive structure of the model, it can be achieved without difficulty to estimate parameters including mean. However, tight conditions on cumulant function, such as analyticity, strict convexity, and steepness, reduce the class of exponential dispersion model. In this work, we present relaxed exponential dispersion model K-LED (Legendre exponential dispersion model with K cumulants). The cumulant function of the proposed model is a convex function of Legendre type having continuous …
Unbiased Estimating Equation on Inverse Divergence and Its Conditions
Media Center Toyohashi University of Technology Toyohashi, Aichi, Japan Email；Department of Computer Science；Engineering Toyohashi University of Technology Toyohashi, Aichi, Japan Email
Many forecasts consist not of point predictions but concern the evolution of quantities. For example, a central bank might predict the interest rates during the next quarter, an epidemiologist might predict trajectories of infection rates, a clinician might predict the behaviour of medical markers over the next day, etc. The situation is further complicated since these forecasts sometimes only concern the approximate "shape of the future evolution" or "order of events". Formally, such forecasts can be seen as probability measures on spaces of equivalence classes of paths modulo time-parametriz…
