Multivariate Rényi divergences characterise betting games with multiple lotteries

• URL: http://arxiv.org/abs/2601.17850v1
• Date: Sun, 25 Jan 2026 14:15:30 GMT
• Title: Multivariate Rényi divergences characterise betting games with multiple lotteries
• Authors: Andrés F. Ducuara, Erkka Haapasalo, Ryo Takakura,
• Abstract summary: We show that the multivariate Rényi divergence $D_underline(vecP_X)$ of probability distributions $vecP_X =(p(0)_X,dots,p(d)_X)$ quantifies the economic-theoretic value that a rational agent assigns to lotteries.<n>In particular, when the odds are fair and the rational agent maximises over all betting strategies, the economic-theoretic value that the agent assigns to lotteries is exactly given by $wmathrmICE_underlineR
• Score: 0.9558392439655014
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We provide an operational interpretation of the multivariate Rényi divergence in terms of economic-theoretic tasks based on betting, risk aversion, and multiple lotteries. We show that the multivariate Rényi divergence $D_{\underlineα}(\vec{P}_X)$ of probability distributions $\vec{P}_X =(p^{(0)}_X,\dots,p^{(d)}_X)$ and real-valued orders $\underlineα = (α_0, \dots, α_d)$ quantifies the economic-theoretic value that a rational agent assigns to $d$ lotteries with odds $o^{(k)}_X \propto (p_X^{(k)})^{-1}$ ($k=1,\dots,d$) on a random event described by $p^{(0)}_X$. In particular, when the odds are fair and the rational agent maximises over all betting strategies, the economic-theoretic value (the isoelastic certainty equivalent) that the agent assigns to the lotteries is exactly given by $w^{\mathrm{ICE}}_{\underline{R}}=\exp[D_{\underlineα}(\vec{P}_X)]$, where $\underline{R}=(R_1,\dots,R_d)$ is a risk-aversion vector with $R_k = 1+α_k/α_0$ being the risk-aversion parameter for lottery $k$. Furthermore, we introduce a new conditional multivariate Rényi divergence that characterises a generalised scenario where the agent uses side information. We prove that this new quantity satisfies a data processing inequality which can be interpreted as the increment in the economic-theoretic value provided by side information; crucially, such a data processing inequality is a consequence of the agent's economic-theoretically consistent risk-averse attitude towards every lottery and vice versa. Finally, we apply these results to the resource theory of informative measurements in general probabilistic theories (GPTs). By establishing quantitative connections between information theory, physics, and economics, our framework provides a novel operational foundation for quantum state betting games with multiple lotteries in the realm of quantum resource theories.

Related papers

• Optimal Unconstrained Self-Distillation in Ridge Regression: Strict Improvements, Precise Asymptotics, and One-Shot Tuning [61.07540493350384]
  Self-distillation (SD) is the process of retraining a student on a mixture of ground-truth and the teacher's own predictions.<n>We show that for any prediction risk, the optimally mixed student improves upon the ridge teacher for every regularization level.<n>We propose a consistent one-shot tuning method to estimate $star$ without grid search, sample splitting, or refitting.
  arXiv  Detail & Related papers  (2026-02-19T17:21:15Z)
• A Relative-Budget Theory for Reinforcement Learning with Verifiable Rewards in Large Language Model Reasoning [48.70183357021465]
  Reinforcement learning (RL) is a dominant paradigm for improving the reasoning abilities of large language models.<n>We propose a emphrelative-budget theory explaining this variation through a single quantity called relative budget $:= H/mathbbE[T]$.<n>We show that $$ determines sample efficiency by controlling reward variance and the likelihood of informative trajectories.
  arXiv  Detail & Related papers  (2026-02-02T01:31:52Z)
• A Theory of Diversity for Random Matrices with Applications to In-Context Learning of Schrödinger Equations [8.997633416528586]
  Given a collection $mathbfA(1), dots, mathbfA(N)$, what is the probability that the centralizer of $mathbfA(1), dots, mathbfA(N)$ is trivial?<n>We provide lower bounds on this probability in terms of the sample size $N$ and the dimension $d$ for several families of random matrices which arise from the discretization of linear Schrdinger operators with random potentials.
  arXiv  Detail & Related papers  (2026-01-18T21:12:54Z)
• PT-Symmetric $SU(2)$-like Random Matrix Ensembles: Invariant Distributions and Spectral Fluctuations [0.0]
  The randomness of the ensemble is endowed by obtaining probability distributions based on symmetry and statistical independence.<n>The degree of level repulsion is a parameter of great interest as it makes a connection to quantum chaos.
  arXiv  Detail & Related papers  (2025-01-11T17:34:09Z)
• Mind the Gap: A Causal Perspective on Bias Amplification in Prediction & Decision-Making [58.06306331390586]
  We introduce the notion of a margin complement, which measures how much a prediction score $S$ changes due to a thresholding operation. We show that under suitable causal assumptions, the influences of $X$ on the prediction score $S$ are equal to the influences of $X$ on the true outcome $Y$.
  arXiv  Detail & Related papers  (2024-05-24T11:22:19Z)
• TIC-TAC: A Framework for Improved Covariance Estimation in Deep Heteroscedastic Regression [109.69084997173196]
  Deepscedastic regression involves jointly optimizing the mean and covariance of the predicted distribution using the negative log-likelihood. Recent works show that this may result in sub-optimal convergence due to the challenges associated with covariance estimation. We study two questions: (1) Does the predicted covariance truly capture the randomness of the predicted mean? Our results show that not only does TIC accurately learn the covariance, it additionally facilitates an improved convergence of the negative log-likelihood.
  arXiv  Detail & Related papers  (2023-10-29T09:54:03Z)
• On Regression in Extreme Regions [0.7974430263940756]
  We establish a statistical learning theoretical framework aimed at extrapolation, or out-of-domain generalization, on the unobserved tails of covariates.<n>We address the stylized problem of nonparametric least squares regression with predictors chosen from a Vapnik-Chervonenkis class.<n>We quantify the predictive performance on tail regions in terms of excess risk, presenting it as a finite sample risk bound with a clear bias-variance decomposition.
  arXiv  Detail & Related papers  (2023-03-06T12:55:38Z)
• Statistical Learning under Heterogeneous Distribution Shift [71.8393170225794]
  Ground-truth predictor is additive $mathbbE[mathbfz mid mathbfx,mathbfy] = f_star(mathbfx) +g_star(mathbfy)$.
  arXiv  Detail & Related papers  (2023-02-27T16:34:21Z)
• Probably Anytime-Safe Stochastic Combinatorial Semi-Bandits [81.60136088841948]
  We propose an algorithm that minimizes the regret over the horizon of time $T$. The proposed algorithm is applicable to domains such as recommendation systems and transportation.
  arXiv  Detail & Related papers  (2023-01-31T03:49:00Z)
• Minimax-Optimal Multi-Agent RL in Zero-Sum Markov Games With a Generative Model [50.38446482252857]
  Two-player zero-sum Markov games are arguably the most basic setting in multi-agent reinforcement learning. We develop a learning algorithm that learns an $varepsilon$-approximate Markov NE policy using $$ widetildeObigg. We derive a refined regret bound for FTRL that makes explicit the role of variance-type quantities.
  arXiv  Detail & Related papers  (2022-08-22T17:24:55Z)
• Risk-averse Contextual Multi-armed Bandit Problem with Linear Payoffs [7.125769932993104]
  We consider the contextual multi-armed bandit problem for linear payoffs under a risk-averse criterion. At each round, contexts are revealed for each arm, and the decision maker chooses one arm to pull and receives the corresponding reward. We apply the Thompson Sampling algorithm for the disjoint model, and provide a comprehensive regret analysis for a variant of the proposed algorithm.
  arXiv  Detail & Related papers  (2022-06-24T18:48:35Z)
• Optimal Sub-Gaussian Mean Estimation in $\mathbb{R}$ [5.457150493905064]
  We present a novel estimator with sub-Gaussian convergence. Our estimator does not require prior knowledge of the variance. Our estimator construction and analysis gives a framework generalizable to other problems.
  arXiv  Detail & Related papers  (2020-11-17T02:47:24Z)
• Distributed Bandits: Probabilistic Communication on $d$-regular Graphs [5.33024001730262]
  We study the decentralized multi-agent multi-armed bandit problem for agents that communicate with probability over a network defined by a $d$-regular graph. We propose a new Upper Confidence Bound (UCB) based algorithm and analyze how agent-based strategies contribute to minimizing group regret.
  arXiv  Detail & Related papers  (2020-11-16T04:53:54Z)
• Reverse Euclidean and Gaussian isoperimetric inequalities for parallel sets with applications [0.0]
  We show that the surface area of an $r$-parallel set in $mathbb Rd$ with volume at most $V$ is upper-bounded by $eTheta(d)V/r$, whereas its Gaussian surface area is upper-bounded by $max(eTheta(d), eTheta(d)/r)$. We also derive a reverse form of the Brunn-Minkowski inequality for $r$-parallel sets, and as an aside a reverse entropy power inequality for Gaussian-smoothed random variables
  arXiv  Detail & Related papers  (2020-06-16T23:58:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.