arXiv subjects

David Ríos Insua

Publications and source records attributed to David Ríos Insua.

18 recordsLinked to original sources

arXiv2026

Shrinkage through multiple identifiability

We propose an empirical Bayes framework for combining estimators obtained from multiple identification functionals associated with the same estimand. We adaptively pool a collection of asymptotically linear estimators, each of which may target a different parameter because of violations of their identification assumptions. Although all estimators are computed from the same sample and are therefore dependent, we show that a working independence construction preserves consistency of the posterior mean under centered heterogeneity. Inference is driven by a latent heterogeneity parameter governing the dispersion of the different identification images. When this parameter is zero, the functionals share a common estimand and we construct frequentist confidence intervals using either sandwich variance estimation or subsampling. When it is positive, the functionals are interpreted as exchangeable draws from a latent population of causal effects, and we construct asymptotically valid Bayesian prediction intervals for the latent target of a new identification functional. These inferential procedures answer different questions, rely on distinct assumptions, and are therefore complementary rather than competing. We illustrate the framework by augmenting evidence from randomized controlled trials with observational studies.

stat.ME↗

arXiv2026

A probabilistic framework for online test-time adaptation

This paper presents a probabilistic framework for online test-time adaptation problems. In them, a model is trained on labeled data but must adapt to unlabeled data at test time under the assumption that training and test distributions potentially differ, that is, there might have been a distributional shift. The framework is based on a state-space modelling architecture from which parameter learning, parameter time evolution, prior tuning, and prediction can be characterized.

stat.ML↗

arXiv2026

A unifying Bayesian framework for adversarial robustness

The vulnerability of machine learning models to adversarial attacks remains a critical societal security challenge. Traditional defenses, such as adversarial training, typically robustify models by minimizing a worst-case loss. These deterministic approaches do not account for uncertainty in the adversary's attack. While stochastic defenses placing a probability distribution on the adversary exist, they often lack statistical rigor and fail to make explicit their underlying assumptions. To resolve these issues, we introduce a formal Bayesian framework that models adversarial uncertainty through a stochastic channel, articulating all probabilistic assumptions. This yields two robustification strategies: a proactive defense enacted during training, aligned with adversarial training, and a reactive defense enacted during operations, aligned with adversarial purification. Several state-of-the-art defenses can be recovered as limiting cases of our model. We empirically validate our methodology, showcasing the benefits of explicitly modeling adversarial uncertainty.

stat.ML↗

arXiv2025

Predictive posteriors under hidden confounding

Predicting outcomes in external domains is challenging due to hidden confounders that potentially influence both predictors and outcomes. Well-established methods frequently rely on stringent assumptions, explicit knowledge about the distribution shift across domains, or bias-inducing regularization schemes to enhance generalization. While recent developments in point prediction under hidden confounding attempt to mitigate these shortcomings, they generally do not provide principled uncertainty quantification. We introduce a Bayesian framework that yields well-calibrated predictive distributions across external domains, supports valid model inference, and achieves posterior contraction rates that improve as the number of observed datasets increases. Simulations and a medical application highlight the remarkable empirical coverage of our approach, nearly unchanged when transitioning from low- to moderate-dimensional settings.

stat.ME↗

arXiv2025

Evasion Attacks Against Bayesian Predictive Models

There is an increasing interest in analyzing the behavior of machine learning systems against adversarial attacks. However, most of the research in adversarial machine learning has focused on studying weaknesses against evasion or poisoning attacks to predictive models in classical setups, with the susceptibility of Bayesian predictive models to attacks remaining underexplored. This paper introduces a general methodology for designing optimal evasion attacks against such models. We investigate two adversarial objectives: perturbing specific point predictions and altering the entire posterior predictive distribution. For both scenarios, we propose novel gradient-based attacks and study their implementation and properties in various computational setups.

stat.ML↗

arXiv2025

Supporting product launching decisions with adversarial risk analysis

In a world of utility-driven marketing, each company acts as an adversary to other contenders, with all having competing interests. A major challenge for companies launching a new product is that, despite testing, flaws in their product can remain, potentially risking a loss in market share. However, delayed launch decisions can lead to losing first-mover advantages. Furthermore, each company generally has incomplete information on the launch strategy and the product quality of competing brands. From a buyer's perspective, along with the price, customers need to make their buying decisions based on noisy signals, e.g.\ regarding the quality of competing brands. This paper proposes how to support product launch decisions by a company in the presence of several competitors and multiple buyers, with the aid of adversarial risk analysis methods. We illustrate applications in two software launch cases that require deciding about timing, pricing, and quality, referring to single and multiple product purchases.

stat.AP↗

arXiv2025

Unsupervised domain adaptation under hidden confounding

We introduce a new predictive mechanism that operates in the presence of hidden confounding across distributionally diverse data sources while ensuring consistent estimation of causal parameters-despite their recognized suboptimality for prediction in the literature. Our method is based on a novel estimand that captures the dependence structure between response noise and covariates, incorporating causal parameters into a generative model that adaptively replicates the conditional distribution of the test environment. Identifiability is achieved under a straightforward, empirically verifiable assumption. Our approach ensures probabilistic alignment with test distributions uniformly across arbitrary interventions, enabling valid predictions without requiring worst-case optimization or assumptions about the strength of perturbations at test time. Through extensive simulations, we demonstrate that our method outperforms state-of-the-art invariance-based and domain adaptation approaches. Additionally, we validate its practical applicability and superior target risk performance on a cardiovascular disease dataset.

math.ST↗

arXiv2025

A decision analysis model for colorectal cancer screening

Background and Objective. With minor differences, most national colorectal cancer (CRC) screening programs in Europe consist of one-size-fits-all aged-based strategies. This paper provides a decision analysis-based approach to personalized CRC screening, supporting decisions concerning whether and which screening method to consider and/or whether a colonoscopy should be administered. Methods. We use an influence diagram which characterizes CRC risk with respect to different variables of interest and includes comfort, costs, complications, and information as decision criteria, the last one assessed through information theory measures. The criteria are integrated with a multi-attribute utility model. Optimal screening policies are then computed. Results. The proposed model is used to support personalized individual screening based on relevant characteristics. It serves to assess existing national screening programs and design new ones. In particular, it suggests replacing current age-based strategies followed in many European countries by more personalized strategies based on the type of model proposed. Additionally, the model facilitates benchmarking of novel screening devices. Conclusions. This work creates a framework supporting personalized CRC screening improving upon current age-based screening strategies.

stat.AP↗

arXiv2024

Colorectal cancer risk mapping through Bayesian Networks

Background and Objective: Only about 14 % of eligible EU citizens finally participate in colorectal cancer (CRC) screening programs despite it being the third most common type of cancer worldwide. The development of CRC risk models can enable predictions to be embedded in decision-support tools facilitating CRC screening and treatment recommendations. This paper develops a predictive model that aids in characterizing CRC risk groups and assessing the influence of a variety of risk factors on the population. Methods: A CRC Bayesian Network is learnt by aggregating extensive expert knowledge and data from an observational study and making use of structure learning algorithms to model the relations between variables. The network is then parametrized to characterize these relations in terms of local probability distributions at each of the nodes. It is finally used to predict the risks of developing CRC together with the uncertainty around such predictions. Results: A graphical CRC risk mapping tool is developed from the model and used to segment the population into risk subgroups according to variables of interest. Furthermore, the network provides insights on the predictive influence of modifiable risk factors such as alcohol consumption and smoking, and medical conditions such as diabetes or hypertension linked to lifestyles that potentially have an impact on an increased risk of developing CRC. Conclusions: CRC is most commonly developed in older individuals. However, some modifiable behavioral factors seem to have a strong predictive influence on its potential risk of development. Modelling these effects facilitates identifying risk groups and targeting influential variables which are subsequently helpful in the design of screening and treatment programs.

stat.AP↗

arXiv2024

Personalized Pricing Decisions Through Adversarial Risk Analysis

Pricing decisions stand out as one of the most critical tasks a company faces, particularly in today's digital economy. As with other business decision-making problems, pricing unfolds in a highly competitive and uncertain environment. Traditional analyses in this area have heavily relied on game theory and its variants. However, an important drawback of these approaches is their reliance on common knowledge assumptions, which are hardly tenable in competitive business domains. This paper introduces an innovative personalized pricing framework designed to assist decision-makers in undertaking pricing decisions amidst competition, considering both buyer's and competitors' preferences. Our approach (i) establishes a coherent framework for modeling competition mitigating common knowledge assumptions; (ii) proposes a principled method to forecast competitors' pricing and customers' purchasing decisions, acknowledging major business uncertainties; and, (iii) encourages structured thinking about the competitors' problems, thus enriching the solution process. To illustrate these properties, in addition to a general pricing template, we outline two specifications - one from the retail domain and a more intricate one from the pension fund domain.

cs.GT↗

arXiv2024

A framework for expected capability sets

This paper addresses decision-aiding problems that involve multiple objectives and uncertain states of the world. Inspired by the capability approach, we focus on cases where a policy maker chooses an act that, combined with a state of the world, leads to a set of choices for citizens. While no preferential information is available to construct importance parameters for the criteria, we can obtain likelihoods for the different states. To effectively support decision-aiding in this context, we propose two procedures that merge the potential set of choices for each state of the world taking into account their respective likelihoods. Our procedures satisfy several fundamental and desirable properties that characterize the outcomes.

cs.CY↗

arXiv2021

Managing driving modes in automated driving systems

Current technologies are unable to produce massively deployable, fully autonomous vehicles that do not require human intervention. Such technological limitations are projected to persist for decades. Therefore, roadway scenarios requiring a driver to regain control of a vehicle, and vice versa, will remain critical to the safe operation of semi-autonomous vehicles for the foreseeable future. Herein, we adopt a comprehensive perspective on this problem taking into account the operational design domain, driver and environment monitoring, trajectory planning, and driver intervention performance assessment. Leveraging decision analysis and Bayesian forecasting, both the support of driving mode management decisions and the issuing of early warnings to the driver are addressed. A statistical modeling framework is created and a suite of algorithms are developed to manage driving modes and issue relevant warnings in accordance with the management by exception principle. The efficacy of these developed methods are then illustrated and examined via a simulated case study.

stat.AP↗

arXiv2021

Data sharing games

Data sharing issues pervade online social and economic environments. To foster social progress, it is important to develop models of the interaction between data producers and consumers that can promote the rise of cooperation between the involved parties. We formalize this interaction as a game, the data sharing game, based on the Iterated Prisoner's Dilemma and deal with it through multi-agent reinforcement learning techniques. We consider several strategies for how the citizens may behave, depending on the degree of centralization sought. Simulations suggest mechanisms for cooperation to take place and, thus, achieve maximum social utility: data consumers should perform some kind of opponent modeling, or a regulator should transfer utility between both players and incentivise them.

cs.GT↗

arXiv2020

Adversarial Risk Analysis (Overview)

Adversarial risk analysis (ARA) is a relatively new area of research that informs decision-making when facing intelligent opponents and uncertain outcomes. It enables an analyst to express her Bayesian beliefs about an opponent's utilities, capabilities, probabilities and the type of strategic calculation that the opponent is using. Within that framework, the analyst then solves the problem from the perspective of the opponent while placing subjective probability distributions on all unknown quantities. This produces a distribution over the actions of the opponent that permits the analyst to maximize her expected utility. This overview covers conceptual, modeling, computational and applied issues in ARA.

cs.GT↗

arXiv2019

Assessing Supply Chain Cyber Risks

Risk assessment is a major challenge for supply chain managers, as it potentially affects business factors such as service costs, supplier competition and customer expectations. The increasing interconnectivity between organisations has put into focus methods for supply chain cyber risk management. We introduce a general approach to support such activity taking into account various techniques of attacking an organisation and its suppliers, as well as the impacts of such attacks. Since data is lacking in many respects, we use structured expert judgment methods to facilitate its implementation. We couple a family of forecasting models to enrich risk monitoring. The approach may be used to set up risk alarms, negotiate service level agreements, rank suppliers and identify insurance needs, among other management possibilities.

stat.ML↗

arXiv2019

Gradient Methods for Solving Stackelberg Games

Stackelberg Games are gaining importance in the last years due to the raise of Adversarial Machine Learning (AML). Within this context, a new paradigm must be faced: in classical game theory, intervening agents were humans whose decisions are generally discrete and low dimensional. In AML, decisions are made by algorithms and are usually continuous and high dimensional, e.g. choosing the weights of a neural network. As closed form solutions for Stackelberg games generally do not exist, it is mandatory to have efficient algorithms to search for numerical solutions. We study two different procedures for solving this type of games using gradient methods. We study time and space scalability of both approaches and discuss in which situation it is more appropriate to use each of them. Finally, we illustrate their use in an adversarial prediction problem.

cs.GT↗

arXiv2019

Adversarial classification: An adversarial risk analysis approach

Classification problems in security settings are usually contemplated as confrontations in which one or more adversaries try to fool a classifier to obtain a benefit. Most approaches to such adversarial classification problems have focused on game theoretical ideas with strong underlying common knowledge assumptions, which are actually not realistic in security domains. We provide an alternative framework to such problem based on adversarial risk analysis, which we illustrate with several examples. Computational and implementation issues are discussed.

stat.ML↗

arXiv2018

Large Scale Automated Forecasting for Monitoring Network Safety and Security

Real time large scale streaming data pose major challenges to forecasting, in particular defying the presence of human experts to perform the corresponding analysis. We present here a class of models and methods used to develop an automated, scalable and versatile system for large scale forecasting oriented towards safety and security monitoring. Our system provides short and long term forecasts and uses them to detect safety and security issues in relation with multiple internet connected devices well in advance they might take place.

stat.AP↗