Search

We present a method for quantitative, in-depth analyses of fairness issues in AI systems with an application to credit scoring. To this aim we use BRIO, a tool for the evaluation of AI systems with respect to social unfairness and, more in general, ethically undesirable behaviours. It features a model-agnostic bias detection module, presented in \cite{DBLP:conf/beware/CoragliaDGGPPQ23}, to which a full-fledged unfairness risk evaluation module is added. As a case study, we focus on the context of credit scoring, analysing the UCI German Credit Dataset \cite{misc_statlog_(german_credit_data)_144}. We apply the BRIO fairness metrics to several, socially sensitive attributes featured in the German Credit Dataset, quantifying fairness across various demographic segments, with the aim of identifying potential sources of bias and discrimination in a credit scoring model. We conclude by combining our results with a revenue analysis.

AI systems are not intrinsically neutral and biases trickle in any type of technological tool. In particular when dealing with people, AI algorithms reflect technical errors originating with mislabeled data. As they feed wrong and discriminatory classifications, perpetuating structural racism and marginalization, these systems are not systematically guarded against bias. In this article we consider the problem of bias in AI systems from the point of view of Information Quality dimensions. We illustrate potential improvements of a bias mitigation tool in gender classification errors, referring to two typically difficult contexts: the classification of non-binary individuals and the classification of transgender individuals. The identification of data quality dimensions to implement in bias mitigation tool may help achieve more fairness. Hence, we propose to consider this issue in terms of completeness, consistency, timeliness and reliability, and offer some theoretical results.

The extensive deployment of probabilistic algorithms has radically changed our perspective on several well-established computational notions. Correctness is probably the most basic one. While a typical probabilistic program cannot be said to compute the correct result, we often have quite strong expectations about the frequency with which it should return certain outputs. In these cases, trust as a generalisation of correctness fares better. One way to understand it is to say that a probabilistic computational process is trustworthy if the frequency of its outputs is compliant with a probability distribution which models its expected behaviour. We present a formal computational framework that formalises this idea. In order to do so, we define a typed lambda-calculus that features operators for conducting experiments at runtime on probabilistic programs and for evaluating whether they compute outputs as determined by a target probability distribution. After proving some fundamental computational properties of the calculus, such as progress and termination, we define a static notion of confidence that allows to prove that our notion of trust behaves correctly with respect to the basic tenets of probability theory.

In this paper we present the probabilistic typed natural deduction calculus TPTND, designed to reason about and derive trustworthiness properties of probabilistic computational processes, like those underlying current AI applications. Derivability in TPTND is interpreted as the process of extracting $n$ samples of possibly complex outputs with a certain frequency from a given categorical distribution. We formalize trust for such outputs as a form of hypothesis testing on the distance between such frequency and the intended probability. The main advantage of the calculus is to render such notion of trustworthiness checkable. We present a computational semantics for the terms over which we reason and then the semantics of TPTND, where logical operators as well as a Trust operator are defined through introduction and elimination rules. We illustrate structural and metatheoretical properties, with particular focus on the ability to establish under which term evolutions and logical rules applications the notion of trustworhtiness can be preserved.

In recent years probabilistic model checking has become an important area of research because of the diffusion of computational systems of stochastic nature. Despite its great success, standard probabilistic model checking suffers the limitation of requiring a sharp specification of the probabilities governing the model behaviour. The theory of imprecise probabilities offers a natural approach to overcome such limitation by a sensitivity analysis with respect to the values of these parameters. However, only extensions based on discrete-time imprecise Markov chains have been considered so far for such a robust approach to model checking. We present a further extension based on imprecise Markov reward models. In particular, we derive efficient algorithms to compute lower and upper bounds of the expected cumulative reward and probabilistic bounded rewards based on existing results for imprecise Markov chains. These ideas are tested on a real case study involving the spend-down costs of geriatric medicine departments., Comment: Forthcoming in the proceedings of ISIPTA 2021 (International Symposium of Imprecise Probability: Theory and Applications)

We introduce a method for the assessment of trust for n-open systems based on a measurement of fidelity and present a prototypic implementation of a complaint architecture. We construct a MAPE loop which monitors the compliance between corresponding figures of interest in cyber- and physical domains; derive measures of the system's trustworthiness; and use them to plan and execute actions aiming at guaranteeing system safety and resilience. We conclude with a view on our future work., Comment: Draft version of a paper accepted for publication in the Proceedings of to ANTIFRAGILE 2015 (2nd International Workshop on Computational Antifragility and Antifragile Engineering, https://sites.google.com/site/antifragile15/)