No description available.

No description available.

No description available.

No description available.

We introduce a new, group fairness-inspired stochastic multi-armed bandit problem in the pure exploration setting. We look at the discrepancy between an arm’s mean reward from a group and the highest mean reward for any arm from that group, and call this the disappointment that group suffers from that arm. We define the optimal arm to be the one that minimizes the maximum disappointment over all groups. This optimal arm addresses one problem with maximin fairness, where the group used to choose the maximin best arm suffers little disappointment regardless of what arm is picked, but another group suffers significantly more disappointment by picking that arm as the best one. The challenge of this problem is that the highest mean reward for a group and the arm that gives that reward are unknown. This means we need to pull arms for multiple goals: to find the optimal arm, and to estimate the highest mean reward of certain groups. This leads to the new adaptive sampling algorithm for best arm identification in the fixed confidence setting called MD-LUCB, or Minimax Disappointment LUCB. We prove bounds on MD-LUCB’s sample complexity and then study its performance with empirical simulations. / Graduate

Yes / The concept of ‘fairness’ has been used to frame political struggles by politicians and activists across the political spectrum. This article looks at its use in the US State of Wisconsin during the ‘Uprising’ – a series of occupations, protests, recall elections and militant direct action that began in 2011. These events were a response to a ‘budget repair bill’ that sought to strip public sector union members of their collective bargaining rights and to apply severe austerity measures within the State. This article suggests that although ‘fairness’ has a certain broad-based and intuitive appeal, its mutability means that it is unlikely to be successful in framing a structural critique that can build and sustain social action. Instead, it argues that framing this conflict as an uprising suggested a more explicit form of resistance that enabled a wider mobilization, and this can best be understood as an example of social (movement) unionism – the extension of traditional work place rights approaches to include broader agendas of social justice, civil rights, immigrant rights and economic justice for non-unionized workers.

The purpose of the current study was to investigate whether the separate contributors to procedural and distributive justice also affected the other form of justice. Previous research investigating these cross over effects of justice contributors had not examined inputs in addition to outcomes and procedures, and had typically assumed outcome level to be equivalent to the equitableness of outcomes. Subjects were 120 undergraduate psychology students. Outcomes, inputs and procedures were manipulated in a laboratory experiment in order to assess their independent and combined effects on distributive and procedural justice perceptions. In contrast to past research, the current study found a weak and inconsistent effect of procedures on distributive justice perceptions. Outcome level had a strong effect on both procedural and distributive justice perceptions. In addition, outcome fairness was found to effect procedural justice perceptions. When procedures were fair, the equitableness of outcomes influenced distributive justice ratings. When procedures were unfair, however, the equitableness of outcomes did not influence distributive justice judgements. Implications for procedural justice conceptualizations, equity theory and organizations are discussed. / Ph. D.

As the Next-Generation (NextG, 5G and beyond) wireless network supports a wider range of services, optimization of resource allocation plays a crucial role in ensuring efficient use of the (limited) available network resources. Note that resource allocation may require knowledge of network parameters (e.g., channel state information and available power level) for package schedule. However, wireless networks operate in an uncertain environment where, in many practical scenarios, these parameters are unknown before decisions are made. In the absence of network parameters, a network controller, who performs resource allocation, may have to make decisions (aimed at optimizing network performance and satisfying users' QoS requirements) while emph{learning}. To that end, this dissertation studies two novel online learning problems that are motivated by autonomous resource management in NextG. Key contributions of the dissertation are two-fold. First, we study reward maximization under uncertainty with fairness constraints, which is motivated by wireless scheduling with Quality of Service constraints (e.g., minimum delivery ratio requirement) under uncertainty. We formulate a framework of combinatorial bandits with fairness constraints and develop a fair learning algorithm that successfully addresses the tradeoff between reward maximization and fairness constraints. This framework can also be applied to several other real-world applications, such as online advertising and crowdsourcing. Second, we consider global reward maximization under uncertainty with distributed biased feedback, which is motivated by the problem of cellular network configuration for optimizing network-level performance (e.g., average user-perceived Quality of Experience). We study both the linear-parameterized and non-parametric global reward functions, which are modeled as distributed linear bandits and kernelized bandits, respectively. For each model, we propose a learning algorithmic framework that can be integrated with different differential privacy models. We show that the proposed algorithms can achieve a near-optimal regret in a communication-efficient manner while protecting users' data privacy ``for free''. Our findings reveal that our developed algorithms outperform the state-of-the-art solutions in terms of the tradeoff among the regret, communication efficiency, and computation complexity. In addition, our proposed models and online learning algorithms can also be applied to several other real-world applications, e.g., dynamic pricing and public policy making, which may be of independent interest to a broader research community. / Doctor of Philosophy / As the Next-Generation (NextG) wireless network supports a wider range of services, optimization of resource allocation plays a crucial role in ensuring efficient use of the (limited) available network resources. Note that resource allocation may require knowledge of network parameters (e.g., channel state information and available power level) for package schedule. However, wireless networks operate in an uncertain environment where, in many practical scenarios, these parameters are unknown before decisions are made. In the absence of network parameters, a network controller, who performs resource allocation, may have to make decisions (aimed at optimizing network performance and satisfying users' QoS requirements) while emph{learning}. To that end, this dissertation studies two novel online learning problems that are motivated by resource allocation in the presence uncertainty in NextG. Key contributions of the dissertation are two-fold. First, we study reward maximization under uncertainty with fairness constraints, which is motivated by wireless scheduling with Quality of Service constraints (e.g., minimum delivery ratio requirement) under uncertainty. We formulate a framework of combinatorial bandits with fairness constraints and develop a fair learning algorithm that successfully addresses the tradeoff between reward maximization and fairness constraints. This framework can also be applied to several other real-world applications, such as online advertising and crowdsourcing. Second, we consider global reward maximization under uncertainty with distributed biased feedback, which is motivated by the problem of cellular network configuration for optimizing network-level performance (e.g., average user-perceived Quality of Experience). We consider both the linear-parameterized and non-parametric (unknown) global reward functions, which are modeled as distributed linear bandits and kernelized bandits, respectively. For each model, we propose a learning algorithmic framework that integrate different privacy models according to different privacy requirements or different scenarios. We show that the proposed algorithms can learn the unknown functions in a communication-efficient manner while protecting users' data privacy ``for free''. Our findings reveal that our developed algorithms outperform the state-of-the-art solutions in terms of the tradeoff among the regret, communication efficiency, and computation complexity. In addition, our proposed models and online learning algorithms can also be applied to several other real-world applications, e.g., dynamic pricing and public policy making, which may be of independent interest to a broader research community.

Yes / Research has suggested a fundamental connection between fairness and well-being at the individual, relational, and societal levels. Mattering is a multidimensional construct consisting of feeling valued by, and adding value to, self and others. Prior studies have attempted to connect mattering to both fairness and a variety of well-being outcomes. Based on these findings, we hypothesize that mattering acts as a mediator between fairness and well-being. This hypothesis was tested through Covariance-Based Structural Equation Modeling (CB-SEM) using multidimensional measures of fairness, mattering, and well-being. Results from a Latent Path Analysis conducted on a representative sample of 1,051U.S. adults provide support to our hypothesis by revealing a strong direct predictive effect of mattering onto well-being and a strong indirect effect of fairness onto well-being through mattering. Results also show that mattering is likely to fully mediate the relationship between fairness and multiple domains of well-being, except in one case, namely, economic well-being. These findings illustrate the value of a focus on mattering to understand the relationship between fairness and well-being and to provide future directions for theory, research, and practice. Theoretical implications for the experience of citizenship and participation, along with cross-cultural considerations, are also discussed. / Erwin and Barbara Mautner Endowed Chair in Community Well-Being at the University of Miami

In organizations, it is inevitable that some business activities might seem unfair to subordinates. Social accounts—the explanations managers give their subordinates for those decisions—are known to be a useful tool for managing subordinates’ fairness concerns. Over three decades of research, we learn that social accounts are effectiveness in improving subordinates’ fairness perceptions and reducing their negative reactions. Yet, we have only limited understanding about how social accounts affect the perceptions and behaviors of managers—those who construct and give them. The purpose of this dissertation is to examine the extent to which constructing accounts affects account-givers’ perceptions and behaviors. Drawing on research in social account and behavioral ethics, a model was developed to test the positive effect of constructing accounts on unethical behavior (direct effect) through moral disengagement and guilt (indirect effect). In respect to account types, it was hypothesized that constructing justifications would lead to higher moral disengagement, less guilt, and more unethical behavior, compared with constructing excuses. Account feedback was hypothesized to moderate the indirect effects of justifications and excuses on unethical behaviors such that account acceptance would strengthen moral disengagement and weaken guilt, and in turn, increase unethical behavior. Two experimental designed studies were conducted to test these hypotheses. In Study 1, utilizing a sample of 128 management students, constructing accounts was found to have a positive effect on unethical behavior (i.e., nepotism) with guilt but not moral disengagement explaining some of the variances in this relation. In contrast to my hypotheses, constructing excuses was found to increase guilt more than constructing justifications. Using a sample of 136 management students, Study 2 generally replicated the results found in Study 1: constructing accounts was found to increase unethical behavior (i.e., dishonesty) through guilt, with excuses having a greater effect. This dissertation concludes with a discussion on contributions, practical implications, limitations, and the direction for future research on social accounts and behavioral ethics. / Ph. D.