Fairness Index Research Articles

Gradual (In)Compatibility of Fairness Criteria

Impossibility results show that important fairness measures (independence, separation, sufficiency) cannot be satisfied at the same time under reasonable assumptions. This paper explores whether we can satisfy and/or improve these fairness measures simultaneously to a certain degree. We introduce information-theoretic formulations of the fairness measures and define degrees of fairness based on these formulations. The information-theoretic formulations suggest unexplored theoretical relations between the three fairness measures. In the experimental part, we use the information-theoretic expressions as regularizers to obtain fairness-regularized predictors for three standard datasets. Our experiments show that a) fairness regularization directly increases fairness measures, in line with existing work, and b) some fairness regularizations indirectly increase other fairness measures, as suggested by our theoretical findings. This establishes that it is possible to increase the degree to which some fairness measures are satisfied at the same time -- some fairness measures are gradually compatible.

Longitudinal Fairness with Censorship

Recent works in artificial intelligence fairness attempt to mitigate discrimination by proposing constrained optimization programs that achieve parity for some fairness statistic. Most assume availability of the class label, which is impractical in many real-world applications such as precision medicine, actuarial analysis and recidivism prediction. Here we consider fairness in longitudinal right-censored environments, where the time to event might be unknown, resulting in censorship of the class label and inapplicability of existing fairness studies. We devise applicable fairness measures, propose a debiasing algorithm, and provide necessary theoretical constructs to bridge fairness with and without censorship for these important and socially-sensitive tasks. Our experiments on four censored datasets confirm the utility of our approach.

Cooperative Multi-Agent Fairness and Equivariant Policies

We study fairness through the lens of cooperative multi-agent learning. Our work is motivated by empirical evidence that naive maximization of team reward yields unfair outcomes for individual team members. To address fairness in multi-agent contexts, we introduce team fairness, a group-based fairness measure for multi-agent learning. We then prove that it is possible to enforce team fairness during policy optimization by transforming the team's joint policy into an equivariant map. We refer to our multi-agent learning strategy as Fairness through Equivariance (Fair-E) and demonstrate its effectiveness empirically. We then introduce Fairness through Equivariance Regularization (Fair-ER) as a soft-constraint version of Fair-E and show that it reaches higher levels of utility than Fair-E and fairer outcomes than non-equivariant policies. Finally, we present novel findings regarding the fairness-utility trade-off in multi-agent settings; showing that the magnitude of the trade-off is dependent on agent skill.

Counterclaims and legitimacy in investment treaty arbitration

Abstract Investor-State dispute resolution (ISDS) is currently facing a crisis of legitimacy, notably because of its perceived asymmetry as well as the difficulties it is faced with when trying to reconcile investor rights with human rights. This article wishes to explore the possibility for respondent States to present counterclaims to investment arbitral tribunals as a way of dealing with these criticisms. Even though some rare awards have entertained them favourably, the majority of arbitral tribunals have declined to hear counterclaims. The author, therefore, examines the obstacles existing in investment case law to the admissibility of counterclaims, both from a procedural as well as from a substantive point of view. The analysis then focuses on the beneficial and adverse consequences that would stem from the hypothetical adoption of a set of rules more favorable to the filing of counterclaims by respondent States in investment arbitration. The essay concludes that such an adoption would be warranted and reinstitute a measure of procedural fairness in a framework that has been suffering from a perception of imbalance and a lack of legitimacy.

Cooperative Bargaining Game-Based Scheduling Model With Variable Multiobjective Weights in a UWB and 5G Embedded Workshop

The IoT technology realizes deep integration of the physical process and production information in a manufacturing workshop through the real-time acquisition of data and seamless interaction of equipment, showing valuable academic and applicable potential in upgrading the workshop environment and eliminating the effect of dynamics. This brings a new opportunity to increase workshop productivity. However, how to realize active perception, dynamic optimization, and real-time regulation of the manufacturing process based on IoT technology is a research hotspot facing technical and theoretical challenges. To address these issues, this study constructs a multiagent-based dynamic scheduling (MADS) workshop IoT architecture by embedding advanced Ultra Wide Band (UWB) and 5th-Generation (5G) communication technologies. Differing from conventional scheduling models, the proposed model optimally assigns processes to machines by designing a resource scheduling agency (RS Agency). The NP-complete of multi-objective optimization DJSP is proved and reduced to a cooperative bargaining game negotiation model (CBGNM) to rationalize the allocation of workshop resources. Then, a multi-objective weights tuning scheme enables effective and efficient responses to workshop exceptions. The performance comparisons with public data of the Kacem 8×8 benchmark verify that the CBGNM has better results than the conventional scheduling methods and reaches maximum optimization of 41.7%, 39.3%, 47.2%, 32.9%, 54.5%, and 38.4% regarding makespan, critical machine workload, total processing time, fairness index, total energy consumption, and comprehensive index, respectively. The numerical experiments show that the game players have no significant coupling with their joint payoff and validate that their bargaining power varies with abnormal events.

Fair and Energy-Efficient Coverage Optimization for UAV Placement Problem in the Cellular Network

Unmanned Aerial Vehicle (UAV) Base Station (BS) placement optimization is an essential operational task to improve the Quality of Service (QoS) in UAV-aided wireless cellular networks. The existing approaches are almost zeroth order methods, and the few first order methods mainly ignore the allocation fairness, computational efficiency, and backhaul constraints. In this paper, we formulate the UAV placement problem as a constrained optimization problem, with the objective of maximizing the fair coverage versus energy consumption while satisfying the backhaul constraints at different time nodes. To guarantee fair QoS allocation, we introduce a novel fairness index to ensure fair communication opportunity and the novel region coverage ratio to avoid excess QoS on covered spots. An accurate and efficient proximal stochastic gradient descent based alternating algorithm that iteratively executes two optimization steps is proposed to optimize the UAV locations, which enables the fast single point-based first order methods to solve the complex problems with constraints. Experiment results manifest that the proposed algorithm performs well both in synthetic data scenario and in real city scenario. Furthermore, the proposed first order algorithm is more efficient than the existing zeroth order algorithm, typically referring to the meta-heuristic method.

A new assessment method of sustainable water resources utilization considering fairness-efficiency-security: A case study of 31 provinces and cities in China

• An evaluation method for SWRU was developed based on water ecological footprint model. • Assessing the performance of SWRU in terms of fairness, efficiency and security. • WSUI was at a relatively unsustainable level in China over the period 2005-2019. • Low water resources efficiency is the most prominent factor limiting SWRU. • Findings can provide scientific reference for improving regional water resources sustainability. Accurate assessment of sustainable water resources utilization (SWRU) is one of the basis for alleviating the growing human-water conflict. However, there are some unavoidable defects in previous research, such as poor holistic considerations, subjectivity in judgment process and limited methodological applicability. Therefore, after determining the triple attributes of SWRU (i.e., fairness, efficiency and security), Gini coefficient, data envelopment analysis (DEA) technique and water ecological footprint (WEF) model were used to establish water resources fairness index (WFI), water resources efficiency index (WEI) and water resources ecological security index (WESI). The three indices were then integrated into water resources sustainable utilization index (WSUI) with Euclidean distance, thus forming the SWRU quantification method system, which can effectively overcome the above shortcomings. Finally, the temporal and spatial characteristics of SWRU during 2005-2019 were diagnosed in 31 provinces and cities of China. Results reveal that: (1) Over half of provinces have a mismatch between population and water distribution, especially in the north, where the contradiction of more people and less water is chronic; (2) The decline in water resources efficiency has not attracted sufficient attention for a long time; (3) WESI in the south was below threshold 1 and water resources ecology is safe, while the northern regions are basically at an insecure level; (4) WSUI multi-year average was within [0.44, 0.51] in China, which has been at relatively unsustainable level, and low water resources efficiency is the most prominent factor limiting SWRU. These findings can provide scientific reference for improving the sustainable utilization of regional water resources.

Energy-Efficient Mobile Sink-Based Intelligent Data Routing Scheme for Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have become an indistinguishable part of any Internet of Things (IoT) based system. In an IoT-based system, WSN monitors different physical objects and forwards the collected data to the base station or gateway node for further processing. Existing routing protocols designed for WSNs are not suitable for IoT-based systems. An intelligent and high-performance routing protocol is essential for real-time data analysis and decision-making. Therefore, this paper proposes a mobile sink-based intelligent data routing scheme for WSNs. Initially, Particle Swarm Optimization is used for optimal cluster formation. After cluster formation, the optimal number of rendezvous points are selected, and a data-gathering tour is planned for the mobile sink. Furthermore, a mobile sink-based intelligent data gathering scheme is proposed to collect the data from the cluster heads. Extensive simulations and mathematical analysis are performed to verify the effectiveness of the proposed scheme. The simulation results are compared with state-of-the-art algorithms in terms of stability period, network lifetime, throughput, average energy consumption, fairness index, data transmission latency, and message overhead.

MmS-TCP: Scalable TCP for Improving Throughput and Fairness in 5G mmWave Networks.

The millimeter-wave (mmWave) band, which can provide data rates of multi-gigabits per second, could play a major role in achieving the throughput goals of 5G networks. However, the high-bandwidth mmWave signal is susceptible to blockage by various obstacles, which results in very large and frequent degradation in the quality of the received signals. TCP, the most representative transport layer protocol, suffers from significant performance degradation due to the very dynamic channel conditions of the mmWave signal. Therefore, in this paper, we propose a congestion control algorithm that guarantees sufficient throughput in 5G mmWave networks and that does not significantly worsen TCP fairness. The proposed algorithm, which is a modification of Scalable TCP (S-TCP) that is designed for high-speed networks, provides a more stable performance than the existing TCP congestion control algorithm in mmWave networks through simple modifications. In various simulation experiments that considered the actual mobile user environment, the proposed mmWave Scalable TCP (mmS-TCP) algorithm demonstrated throughput up to 2.4 times higher than CUBIC TCP in single flow evaluation, and the inter-protocol fairness index when competing with CUBIC flow significantly improved from 0.819 of S-TCP to 0.9733. Moreover, the mmS-TCP algorithm reduced the number of duplicated ACKs by 1/4 compared with S-TCP, and it improved the average total throughput and intra-protocol fairness simultaneously.

An integrated AHP‐ELECTRE and deep reinforcement learning methods for handover performance optimization in an LTE‐A networks

AbstractEmblematic growth of telecommunications networks has also produced high data rates and low user equipment (UE) latency. Also, the third generation partnership project launched long term evolution (LTE) and LTE advance (A), which supports different types of network traffic including video, voice, and so on. A handover is needed to provide semantic interloper ability is the process to move the UE power from serving evolved node B (eNB) to the subsequent eNB without interruption. Consequently, high quality of service (QoS), reliability relies on a smooth execution. If eNBs cannot be selected from available eNBs in an optimized manner to transfer during UE mobility, the criteria for quality of service can be violated. In order to accommodate QoS requires to be more configured for smooth synchronization, eNB range, and trigger points. This article aims to automate handover by means of sophisticated program based on the analytical hierarchy process (AHP) with ELECTRE and deep reinforcement (DR) learning methods. Network performance is scrutinized and simulation is utilized to visualize the proposed scheme. The experiments demonstrated that the proposed system helps to reduce the handover failure rate and handover ping‐pong. Thus, the proposed AHP‐ELECTRE‐DR‐learning method has achieved better results in delay, throughput, fairness index, and packet loss than the existing works like AHP‐TOPSIS‐Q‐learning for handover optimization in LTE‐A, auto tuning optimization for reducing frequent handovers and handover failure in LTE‐A, and particle swarm optimization with adaptive neuro‐fuzzy inference system for LTE and LTE‐A networks, respectively.

Evaluation of the reproducibility and confounders of fair ASL placenta perfusion measurement in normal pregnancies

Arterial spin labeling (ASL) is a very promising non-invasive MRI technique that quantifies placental perfusion using arterial blood water as an endogenous tracer, but lacks reproducibility and standardization, which limits its use in clinical practice. The objective was to standardize and identify confounders of FAIR ASL in placental perfusion estimation. This was a prospective study at Necker Hospital ( NCT04142606 ). Fetal weight estimation and Doppler measurements were performed, followed by MRI on a 1.5 T scanner. A T2-weighted axial FIESTA placenta-centered slice was performed as a reference. Perfusion data were acquired with a 10 mm axial single slice FAIR-SSFSE, repeated 3 times with variable selective inversion thickness (3, 5 and 7 cm). To assess reproducibility, the sequences were performed twice and perfusion was measured on the whole placenta and in 4 regions of interest (ROI) by two operators. We also studied the impact of maternal age, gestational age at examination or placental location on blood flow measurements. Sixty-two patients with normal pregnancy were included. The respective mean blood flow for the 3, 5 and 7-cm-thick inversion was 238.3 ± 68.0 mL/100 g/min, 156.4 ± 46.5 mL/100 g/min and 121.6 ± 39.7 mL/100 g/min respectively (ANOVA, P < 0.01) this demonstrates the importance of the methodology in this evaluation. The thicker section appeared to be more representative of placental perfusion in agreement with data in the literature. However we did not observe a significant difference when averaging the perfusion of the 4 ROI or the whole placenta ( P = 0.21, P = 0.56 and P = 0.85 respectively). The sequences were repeated twice in 10 patients and were reproducible. There was no significant difference according to maternal age, gestational age at examination and placental location (anterior or posterior). This study evaluated the reproducibility of FAIR ASL for measuring placental perfusion and showed that the selective slab inversion thickness has a significant impact on blood flow measurements. Other potential confounders need to be investigated to enable the application of FAIR ASL in clinical practice. A potential future application of ASL would be to detect placental insufficiency.

Perceived fairness as main determinant of patients' satisfaction with care during psychiatric hospitalisation: An observational study

IntroductionPatient satisfaction with care is widely recognized as one of the most important indicator of quality in mental health care. It can impact several treatment outcomes, such as treatment adherence and engagement with services. At the same time, as an outcome in itself, satisfaction with care is also affected by several factors, first and foremost by being coerced. The main aim of this study was to test if perceiving treatment pressures as fair and effective could positively impact patient satisfaction, even more than formal coercive measures. MethodsGlobally, 133 voluntary and involuntary inpatients were interviewed. Socio-demographic and clinical characteristics, including history of previous experiences of formal coercion and legal status of the hospitalisation, were collected through a structured questionnaire and medical charts. The participants were also asked to complete the Index of Fairness and Index of Effectiveness tools as well as a structured questionnaire on satisfaction with care. Simple and multiple linear regressions were performed. ResultsAlthough several factors were found to affect satisfaction with care when taken independently, perceived fairness was the stronger predictor of both satisfaction with treatment (β =.234; p = .022) and satisfaction with decision-making involvement (β =.360; p < .001) when controlling for confounders. ConclusionsOur results point to the paramount importance of developing and implementing interventions that promote procedural fairness in psychiatric treatment and thereby improve patient satisfaction while reducing the risk of disengagement with care.

User relay‐aided energy‐efficient resource allocation in heterogeneous cloud radio access network

AbstractIn this article, we consider a user relay‐aided resource allocation problem for femtocell access points (FAPs) and macrocells to improve energy efficiency in heterogeneous cloud radio access networks. The formulated optimization problem is NP‐hard and mixed‐integer non‐linear programming due to the incredibly complex relation between different optimization variables and their impacts on each other. To solve this non‐convex and NP‐hard optimization problem in an effective and distributed method, we propose an efficient three‐step iterative algorithm. The solution approaches are derived from successive convex approximation and complementary geometric programming by considering maximum power, front‐haul link and baseband unit (BBU) capacity, Jain's fairness index, and quality of service constraints of macro and femtocell users. The simulation results depict that the proposed algorithm can effectively decrease the energy consumption of BBUs by switching off under‐utilized BBUs and offloading the traffic from remote radio heads and macro‐cell to FAPs.

Wireless Lan Performance Enhancement Using Double Deep Q-Networks

Due to the exponential growth in the use of Wi-Fi networks, it is necessary to study its usage pattern in dense environments for which the legacy IEEE 802.11 MAC (Medium Access Control) protocol was not specially designed. Although 802.11ax aims to improve Wi-Fi performance in dense scenarios due to modifications in the physical layer (PHY), however, MAC layer operations remain unchanged, and are not capable enough to provide stable performance in dense scenarios. Potential applications of Deep Learning (DL) to Media Access Control (MAC) layer of WLAN has now been recognized due to their unique features. Deep Reinforcement Learning (DRL) is a technique focused on behavioral sensitivity and control philosophy. In this paper, we have proposed an algorithm for setting optimal contention window (CW) under different network conditions called DRL-based Contention Window Optimization (DCWO). The proposed algorithm operates in three steps. In the initial step, Wi-Fi is being controlled by the 802.11 standards. In the second step, the agent makes the decisions concerning the value of CW after the TRAIN procedure for the proposed algorithm. The final phase begins after the training, defined by a time duration specified by the user. Now, the agent is fully trained, and no updates will be no longer received. Now the CW is updated via the OPTIMIZE process of DCWO. We have selected total network throughput, instantaneous network throughput, fairness index, and cumulative reward, and compared our proposed scheme DCWO with the Centralized Contention window Optimization with DRL (CCOD). Simulation results show that DCWO with Double Deep Q-Networks (DDQN) performs better than CCOD with (i) Deep Deterministic Policy Gradient (DDPG) and (ii) Deep Q-Network (DQN). More specifically, DCWO with DDQN gives on average 28% and 23% higher network throughput than CCOD in static and dynamic scenarios. Whereas in terms of instantaneous network throughput DCWO gives around 10% better results than the CCOD. DCWO achieves almost near to optimal fairness in static scenarios and better than DQN and DDPG with CCOD in dynamic scenarios. Similarly, while the cumulative reward achieved by DCWO is almost the same with CCOD with DDPG, the uptrend of DCWO is still encouraging.

PFPS: Priority-First Packet Scheduler for IEEE 802.15.4 Heterogeneous Wireless Sensor Networks

This paper presents priority-first packet scheduling approach for heterogeneous traffic flows in low data rate heterogeneous wireless sensor networks (HWSNs). A delay sensitive or emergency event occurrence demands the data delivery on the priority basis over regular monitoring sensing applications. In addition, handling sudden multi-event data and achieving their reliability requirements distinctly becomes the challenge and necessity in the critical situations. To address this problem, this paper presents distributed approach of managing data transmission for simultaneous traffic flows over multi-hop topology, which reduces the load of a sink node; and helps to make a life of the network prolong. For this reason, heterogeneous traffic flows algorithm (CHTF) algorithm classifies the each incoming packets either from source nodes or downstream hop node based on the packet priority and stores them into the respective queues. The PFPS-EDF and PFPS-FCFS algorithms present scheduling for each data packets using priority weight. Furthermore, reporting rate is timely updated based on the queue level considering their fairness index and processing rate. The reported work in this paper is validated in ns2 (ns2.32 allinone) simulator by putting the network into each distinct cases for validation of presented work and real time TestBed. The protocol evaluation presents that the distributed queue-based PFPS scheduling mechanism works efficiently using CSMA/CA MAC protocol of the IEEE 802.15.4 sensor networks.

Congestion-free routing based on a hybrid meta-heuristic algorithm to provide an effective routing protocol by analyzing the significant impacts of QoS parameters in a dynamic VANET environment

Abstract A definitive form of Mobile Ad-hoc Network (MANET) is the Vehicular Ad-hoc Network (VANET), which delivers data communication using wireless transmission with the collection of vehicular nodes in a vehicular environment. The main goal of this VANET is to improve the Quality of Service (QoS) which is road safety, environmental kindliness, and logistics in intelligent transport. Moreover, mobility and routing play an important challenge for information interchange to the process in the VANET environment. In developing countries, smart cities are encountering problematic traffic congestion. Furthermore, in cluster communications, stable cluster formation and maintenance with the guarantying QoS has always remained a great challenge. In order to overcome the above-mentioned issue, this article proposes a congestion-free routing process ensuring QOS parameters. The main objective of the proposed scheme is to analyze the impact of QOS parameters while choosing the best routing path in Vehicular communication. The scheme also introduces the Hybrid-Harris Hawks Black Widow optimization-based VANET (H-HHBW-VANET) algorithm to improve QoS for routing between nodes. This algorithm differs from other algorithms in terms of optimal solution generation, which makes it more efficient. The proposed work is implemented in the working platform of NS2 to analyze the parameters such as Drop, Delay, Energy Consumption, Throughput, and Fairness Index. The simulation results provide better QOS service in terms of low energy consumption with higher throughput and low delay.

Just Machines

AbstractA number of findings in the field of machine learning have given rise to questions about what it means for automated scoring- or decision-making systems to be fair. One center of gravity in this discussion is whether such systems ought to satisfy classification parity (which requires parity in predictive performance across groups, defined by protected attributes) or calibration (which requires similar predictions to have similar meanings across groups, defined by protected attributes). Central to this discussion are impossibility results, which show that classification parity and calibration are often incompatible. This paper aims to argue that classification parity, calibration, and a newer, interesting measure called counterfactual fairness are unsatisfactory measures of fairness, offer a general diagnosis of the failure of these measures, and sketch an alternative approach to understanding fairness in machine learning.

Search-based fairness testing for regression-based machine learning systems

ContextMachine learning (ML) software systems are permeating many aspects of our life, such as healthcare, transportation, banking, and recruitment. These systems are trained with data that is often biased, resulting in biased behaviour. To address this issue, fairness testing approaches have been proposed to test ML systems for fairness, which predominantly focus on assessing classification-based ML systems. These methods are not applicable to regression-based systems, for example, they do not quantify the magnitude of the disparity in predicted outcomes, which we identify as important in the context of regression-based ML systems.Method:We conduct this study as design science research. We identify the problem instance in the context of emergency department (ED) wait-time prediction. In this paper, we develop an effective and efficient fairness testing approach to evaluate the fairness of regression-based ML systems. We propose fairness degree, which is a new fairness measure for regression-based ML systems, and a novel search-based fairness testing (SBFT) approach for testing regression-based machine learning systems. We apply the proposed solutions to ED wait-time prediction software.Results:We experimentally evaluate the effectiveness and efficiency of the proposed approach with ML systems trained on real observational data from the healthcare domain. We demonstrate that SBFT significantly outperforms existing fairness testing approaches, with up to 111% and 190% increase in effectiveness and efficiency of SBFT compared to the best performing existing approaches.Conclusion:These findings indicate that our novel fairness measure and the new approach for fairness testing of regression-based ML systems can identify the degree of fairness in predictions, which can help software teams to make data-informed decisions about whether such software systems are ready to deploy. The scientific knowledge gained from our work can be phrased as a technological rule; to measure the fairness of the regression-based ML systems in the context of emergency department wait-time prediction use fairness degree and search-based techniques to approximate it.

Does trait interpersonal fairness moderate situational influence on fairness behavior?

Although fairness is a key moral trait, limited research focuses on participants' observed fairness behavior because moral traits are generally measured through self-report. This experiment focused on day-to-day interpersonal fairness rather than impersonal justice, and fairness was assessed as observed behavior. The experiment investigated whether a self-reported fairness trait would moderate a situational influence on observed fairness behavior, such that individuals with a stronger fairness trait would be less affected by a situational influence than those with a weaker fairness trait. We used an iterated resource game in which participants could withdraw resources as they chose, and we manipulated the number of resources bogus players withdrew. The number of resources participants withdrew was the behavioral measure of fairness. Results confirmed the expected moderation of the unfairness manipulation by a fairness trait on observed behavior. Those reporting a stronger fairness trait were unaffected by the manipulation, whereas those reporting a weaker fairness trait were more strongly influenced.

FAIR: Fairness‐aware information retrieval evaluation

AbstractWith the emerging needs of creating fairness‐aware solutions for search and recommendation systems, a daunting challenge exists of evaluating such solutions. While many of the traditional information retrieval (IR) metrics can capture the relevance, diversity, and novelty for the utility with respect to users, they are not suitable for inferring whether the presented results are fair from the perspective of responsible information exposure. On the other hand, existing fairness metrics do not account for user utility or do not measure it adequately. To address this problem, we propose a new metric called FAIR. By unifying standard IR metrics and fairness measures into an integrated metric, this metric offers a new perspective for evaluating fairness‐aware ranking results. Based on this metric, we developed an effective ranking algorithm that jointly optimized user utility and fairness. The experimental results showed that our FAIR metric could highlight results with good user utility and fair information exposure. We showed how FAIR related to a set of existing utility and fairness metrics and demonstrated the effectiveness of our FAIR‐based algorithm. We believe our work opens up a new direction of pursuing a metric for evaluating and implementing the FAIR systems.