On Meritocracy in Optimal Set Selection

Supervised learning models have been increasingly used for making decisions about individuals in applications such as hiring, lending, and college admission. These models may inherit pre-existing biases from training datasets and discriminate against protected attributes (e.g., race or gender). In addition to unfairness, privacy concerns also arise when the use of models reveals sensitive personal information. Among various privacy notions, differential privacy has become popular in recent years. In this work, we study the possibility of using a differentially private exponential mechanism as a post-processing step to improve both fairness and privacy of supervised learning models. Unlike many existing works, we consider a scenario where a supervised model is used to select a limited number of applicants as the number of available positions is limited. This assumption is well-suited for various scenarios, such as job application and college admission. We use ``equal opportunity'' as the fairness notion and show that the exponential mechanisms can make the decision-making process perfectly fair. Moreover, the experiments on real-world datasets show that the exponential mechanism can improve both privacy and fairness, with a slight decrease in accuracy compared to the model without post-processing.

The expansion of state-mandated tests in the 1990s and the testing requirements of the No Child Left Behind Act have supplied researchers with an abundance of data on test scores that can be used as measures of school quality.This paper uses the state-mandated test scores for 5 th grade and 11 th grade in Montgomery County, Pennsylvania, to examine three issues about the capitalization of school quality into house prices: (1) At what level do prospective home buyers evaluate the quality of local public education-at the district level or the level of the neighborhood school?(2) After accounting for student achievement as reflected in test scores, are other aspects of the local public school system, such as class size or expenditures, capitalized into the value of a house?(3) Are the positive results we get for the capitalization of school quality into house prices due simply to the correlation between high test scores and other desirable neighborhood characteristics?The results of our investigation suggest that to home buyers some test-score averages are significantly better indicators of the quality of the local public school system than others.In particular, home buyers seem to evaluate the quality of public education at the district level rather than at the level of the local school.Class size at the high-school level has some independent effect on house prices, but not class size at the elementary school level.And once we account for student achievement, expenditures per pupil have no further effect on house prices.Finally, restricting our sample to similar neighborhoods along school district boundaries confirms our earlier results for high school test scores but not for elementary school scores.

Conventional and contemporary models of travel choice make the restrictive assumption of homogeneity in decision rules. Recent literature has shown empirical evidence for potential heterogeneity in decision rules with regard to utility maximization and regret minimization. Notwithstanding these advances in modeling decision rules, behavioral understanding in the differences in these alternative decision rule segments has not been sufficiently understood. Moreover, the factors which influence the choice of these decision rules have not received significant attention. This study proposes a framework which considers decision makers to have both utility maximizing and regret minimizing tendencies. The variation in these tendencies across decision makers renders the framework heterogeneous. A heterogeneous decision rule model is developed assuming the decision rule adopted to be a latent construct. The study characterizes the regret minimizing and utility maximizing segments based on average values of the segmental attributes. The empirical findings show evidence to confirm that utility maximizers tend to be predominantly captive to personal vehicle usage while regret minimizers might be non-captive to any particular mode. The nature and extent of influence of factors affecting the choice of decision rule is also examined.

R. Lucas's Research Program: Far Too Narrow with Too Many Special Assumptions

On the Ellsberg Paradox and its Extension by Machina

Fairness is a convenient and popular tool when modelling and specifying concurrent systems, many fairness notions exist in the literature. However, there was no fully satisfactory abstract characterisation of fairness. In order to show fairness definitions are reasonable, a probabilistic framework for fairness notions was proposed, where concurrent systems were modelled inductively in the theorem prover Isabelle/HOL, and a general fairness notion was defined. It was shown that the set of fair infinite executions of concurrent systems has probability 1. Finally, the usage of this framework was argued. All formalisations and proofs have been carried out with Isabelle/HOL/Isar.

Information measures arise in many disciplines, including forecasting (where scoring rules are used to provide incentives for probability estimation), signal processing (where information gain is measured in physical units of relative entropy), decision analysis (where new information can lead to improved decisions), and finance (where investors optimize portfolios based on their private information and risk preferences). In this paper, we generalize the two most commonly used parametric families of scoring rules and demonstrate their relation to well-known generalized entropies and utility functions, shedding new light on the characteristics of alternative scoring rules as well as duality relationships between utility maximization and entropy minimization. In particular, we show that weighted forms of the pseudospherical and power scoring rules correspond exactly to measures of relative entropy (divergence) with convenient properties, and they also correspond exactly to the solutions of expected utility maximization problems in which a risk-averse decision maker whose utility function belongs to the linear-risk-tolerance family interacts with a risk-neutral betting opponent or a complete market for contingent claims in either a one-period or a two-period setting. When the market is incomplete, the corresponding problems of maximizing linear-risk-tolerance utility with the risk-tolerance coefficient β are the duals of the problems of minimizing the pseudospherical or power divergence of order β between the decision maker's subjective probability distribution and the set of risk-neutral distributions that support asset prices.

Language Models (LMs) have demonstrated exceptional performance across various Natural Language Processing (NLP) tasks. Despite these advancements, LMs can inherit and amplify societal biases related to sensitive attributes such as gender and race, limiting their adoption in real‐world applications. Therefore, fairness has been extensively explored in LMs, leading to the proposal of various fairness notions. However, the lack of clear agreement on which fairness definition to apply in specific contexts and the complexity of understanding the distinctions between these definitions can create confusion and impede further progress. To this end, this paper proposes a systematic survey that clarifies the definitions of fairness as they apply to LMs. Specifically, we begin with a brief introduction to LMs and fairness in LMs, followed by a comprehensive, up‐to‐date overview of existing fairness notions in LMs and the introduction of a novel taxonomy that categorizes these concepts based on their transformer architecture: encoder‐only, decoder‐only, and encoder‐decoder LMs. We further illustrate each definition through experiments, showcasing their practical implications and outcomes. Finally, we discuss current research challenges and open questions, aiming to foster innovative ideas and advance the field. The repository is publicly available online at https://github.com/vanbanTruong/Fairness‐in‐Large‐Language‐Models/tree/main/definitions . This article is categorized under: Commercial, Legal, and Ethical Issues > Fairness in Data Mining Commercial, Legal, and Ethical Issues > Social Considerations Technologies > Artificial Intelligence .

The connection between economics and ethics is as old as economics itself, and central to both disciplines. It is an issue that has recently attracted much interest from economists and philosophers. The connection is, in part, a result of the desire of economists to make policy prescriptions, which clearly require some normative criteria. More deeply, much economic theory is founded on the assumption of utility maximization, thereby creating an immediate connection between the foundations of economics and the philosophical literature on utilitarianism and reasons for action. In fact, some influential contemporary approaches to ethics advocate decision-theoretic or game-theoretic foundations of some sort for moral principles, while several economic theorists are now prepared to take into account the ethical dimensions of rational decisions. As a result, it appears that economics and ethics are somehow inextricably linked through theories of rational decision-making. Most of us would probably find it disturbing to concede that there are contradictions between the prescriptions of rationality and the requirements of moral `rightness'. The essays included in the present volume provide a detailed analysis of the connections between ethics and economics as viewed from several different - sometimes conflicting - perspectives. This book, the outcome of a joint meeting of philosophers and economists, has three main themes: the validity of utilitarianism much used by economists, the notion of fairness and equity, and the coherence of the rationality postulate of economics. The book does not reach any final conclusions, but it greatly illuminates the exact areas of possible disagreement and indeed the open-ended nature of ethical reasoning. There is much that economists, and especially welfare economists, can learn from these papers - not least circumspection.

Resource allocation problems are a fundamental domain in which to evaluate the fairness properties of algorithms. The trade-offs between fairness and utilization have a long history in this domain. A recent line of work has considered fairness questions for resource allocation when the demands for the resource are distributed across multiple groups and drawn from probability distributions. In such cases, a natural fairness requirement is that individuals from different groups should have (approximately) equal probabilities of receiving the resource. A largely open question in this area has been to bound the gap between the maximum possible utilization of the resource and the maximum possible utilization subject to this fairness condition. Here, we obtain some of the first provable upper bounds on this gap. We obtain an upper bound for arbitrary distributions, as well as much stronger upper bounds for specific families of distributions that are typically used to model levels of demand. In particular, we find - somewhat surprisingly - that there are natural families of distributions (including Exponential and Weibull) for which the gap is non-existent: it is possible to simultaneously achieve maximum utilization and the given notion of fairness. Finally, we show that for power-law distributions, there is a non-trivial gap between the solutions, but this gap can be bounded by a constant factor independent of the parameters of the distribution.

Muth’s Rational Expectations Hypothesis, That “…The Subjective Probability Distributions Are Distributed Around an Objective Probability Distribution, for a given Information Set…” is an Oxymoron. It Can Be Corrected by Amending It to Read “…The Subjective Probability Distributions Are Distributed Around a Subjective Consensus Probability Distribution, for a given Information Set…”

: The main interest is in investigating several kinds of implications of stochastic programming models and their uses to provide several types of dialogues between formal models and the decision maker himself under risk-taking situations. First, some stochastic programming models which are often discussed in the literature are presented and parametric evaluations using dual evaluators are attempted. Second, we see a close relationship between these models and the mean-standard deviation analysis mainly developed in the context of portfolio selection theory. Third, we explore utility implications of these models in connection with the expected utility maximization principle. Finally, attempts are made to impute the value(s) of key parameters in each of these models and some other utility models from the actual decision made in the mean-standard devication analysis of framework. (Author)

Objectives: Discussions of fairness in criminal justice risk assessments typically lack conceptual precision. Rhetoric too often substitutes for careful analysis. In this article, we seek to clarify the trade-offs between different kinds of fairness and between fairness and accuracy. Methods: We draw on the existing literatures in criminology, computer science, and statistics to provide an integrated examination of fairness and accuracy in criminal justice risk assessments. We also provide an empirical illustration using data from arraignments. Results: We show that there are at least six kinds of fairness, some of which are incompatible with one another and with accuracy. Conclusions: Except in trivial cases, it is impossible to maximize accuracy and fairness at the same time and impossible simultaneously to satisfy all kinds of fairness. In practice, a major complication is different base rates across different legally protected groups. There is a need to consider challenging trade-offs. These lessons apply to applications well beyond criminology where assessments of risk can be used by decision makers. Examples include mortgage lending, employment, college admissions, child welfare, and medical diagnoses.

Automated decision-making systems are massively engaged in different types of businesses, including customer-oriented sectors, and bring countless achievements in persuading customers with more personalized experiences. However, it was observed that the decisions made by the algorithms could bring unfairness to a person or a group of people, according to recent studies. Thus, algorithmic fairness has become a spotlight research area, and defining a concrete version of fairness notions has also become significant research. In existing literature, there are more than 21 definitions of algorithmic fairness. Many studies have shown that each notion has an incompatibility problem, and it is still necessary to make those notions more adaptable to the legal and social principles of the desired sectors. Yet, the constraints of algorithmic fairness for customer-oriented areas have not been thoroughly studied. This motivates us to work on a systematic literature review to investigate the sectors concerned about algorithmic fairness as a significant matter when using machine-based decision-making systems, what are the well-applied algorithmic fairness notions, and why they can or cannot be directly applicable to the customer-oriented sectors, what are the possible algorithmic fairness constraints for the customer-oriented sectors. By applying the standard guidelines of systematic literature review, we explored 65 prominent articles thoroughly. The findings show 43 different ways of algorithmic fairness notions in the varieties of domains. We also identified the three important perspectives to be considered for enhancing algorithmic fairness notions in the customer-oriented sectors.

Laboratory studies have demonstrated beneficial effects of making comparisons on children's analogical reasoning skills. We extend this finding to an observational dataset comprising 42 children. The prevalence of specific comparisons, which identify a feature of similarity or difference, in children's spontaneous speech from 14-58 months is associated with higher scores in tests of verbal and non-verbal analogy in 6th grade. We test two pre-registered hypotheses about how parents influence children's production of specific comparisons: 1) via modelling, where parents produce specific comparisons during the sessions prior to child onset of this behaviour; 2) via responsiveness, where parents respond to their children's earliest specific comparisons in variably engaged ways. We do not find that parent modelling or responsiveness predicts children's production of specific comparisons. However, one of our pre-registered control analyses suggests that parents' global comparisons-comparisons that do not identify a specific feature of similarity or difference-may bootstrap children's later production of specific comparisons, controlling for parent IQ. We present exploratory analyses following up on this finding and suggest avenues for future confirmatory research. The results illuminate a potential route by which parents' behaviour may influence children's early spontaneous comparisons and potentially their later analogical reasoning skills.