Simultaneous hypothesis testing for comparing many functional means
Authors:
Colin Decker,
Dehan Kong,
Stanislav Volgushev
Abstract:
Data with multiple functional recordings at each observational unit are increasingly common in various fields including medical imaging and environmental sciences. To conduct inference for such observations, we develop a paired two-sample test that allows to simultaneously compare the means of many functional observations while maintaining family-wise error rate control. We explicitly allow the nu…
▽ More
Data with multiple functional recordings at each observational unit are increasingly common in various fields including medical imaging and environmental sciences. To conduct inference for such observations, we develop a paired two-sample test that allows to simultaneously compare the means of many functional observations while maintaining family-wise error rate control. We explicitly allow the number of functional recordings to increase, potentially much faster than the sample size. Our test is fully functional and does not rely on dimension reduction or functional PCA type approaches or the choice of tuning parameters. To provide a theoretical justification for the proposed procedure, we develop a number of new anti-concentration and Gaussian approximation results for maxima of $L^2$ statistics which might be of independent interest. The methodology is illustrated on the task-related cortical surface functional magnetic resonance imaging data from Human Connectome Project.
△ Less
Submitted 13 June, 2025;
originally announced June 2025.
Robust Q-Learning for finite ambiguity sets
Authors:
Cécile Decker,
Julian Sester
Abstract:
In this paper we propose a novel $Q$-learning algorithm allowing to solve distributionally robust Markov decision problems for which the ambiguity set of probability measures can be chosen arbitrarily as long as it comprises only a finite amount of measures. Therefore, our approach goes beyond the well-studied cases involving ambiguity sets of balls around some reference measure with the distance…
▽ More
In this paper we propose a novel $Q$-learning algorithm allowing to solve distributionally robust Markov decision problems for which the ambiguity set of probability measures can be chosen arbitrarily as long as it comprises only a finite amount of measures. Therefore, our approach goes beyond the well-studied cases involving ambiguity sets of balls around some reference measure with the distance to reference measure being measured with respect to the Wasserstein distance or the Kullback--Leibler divergence. Hence, our approach allows the applicant to create ambiguity sets better tailored to her needs and to solve the associated robust Markov decision problem via a $Q$-learning algorithm whose convergence is guaranteed by our main result. Moreover, we showcase in several numerical experiments the tractability of our approach.
△ Less
Submitted 15 February, 2025; v1 submitted 5 July, 2024;
originally announced July 2024.
Unique equilibria and substitution effects in a stochastic model of the marriage market
Authors:
Colin Decker,
Elliott H. Lieb,
Robert J. McCann,
Benjamin K. Stephens
Abstract:
Choo-Siow (2006) proposed a model for the marriage market which allows for random identically distributed noise in the preferences of each of the participants. The randomness is McFadden-type, which permits an explicit resolution of the equilibrium preference probabilities. The purpose of this note is to prove uniqueness of the resulting equilibrium marriage distribution, and find a representation…
▽ More
Choo-Siow (2006) proposed a model for the marriage market which allows for random identically distributed noise in the preferences of each of the participants. The randomness is McFadden-type, which permits an explicit resolution of the equilibrium preference probabilities. The purpose of this note is to prove uniqueness of the resulting equilibrium marriage distribution, and find a representation of it in closed form. This allows us to derive smooth dependence of this distribution on exogenous preference and population parameters, and establish sign, symmetry, and size of the various substitution effects, facilitating comparative statics. For example, we show that an increase in the population of men of any given type in this model leads to an increase in single men of each type, and a decrease in single women of each type. We show that an increase in the number of men of a given type increases the equilibrium transfer paid by such men to their spouses, and also increases the percentage of men of that type who choose to remain unmarried. The verification of such properties helps to substantiate the validity of the model. Moreover, we make unexpected predictions which could be tested: the percentage change of type $i$ unmarrieds with respect to fluctuations in the total number of type $j$ men or women turns out to form a symmetric positive-definite matrix $r_{ij}=r_{ji}$ in this model, and thus to satisfy bounds like $|r_{ij}| \le (r_{ii}r_{jj})^{1/2}$. Along the way, we give a new proof for the existence of an equilibrium, based on a strictly convex variational principle and a simple estimate, rather than a fixed point theorem. Fixed point approaches to the existence part of our result have been explored by others \cite{CSS} \cite{Dag} \cite{Fox}, but are much more complicated and yield neither uniqueness, nor comparative statics, nor an explicit representation of the solution.
△ Less
Submitted 5 September, 2011; v1 submitted 8 December, 2010;
originally announced December 2010.
