Distribution Constraints Research Articles

Fairness of recommender systems in the recruitment domain: an analysis from technical and legal perspectives.

Recommender systems (RSs) have become an integral part of the hiring process, be it via job advertisement ranking systems (job recommenders) for the potential employee or candidate ranking systems (candidate recommenders) for the employer. As seen in other domains, RSs are prone to harmful biases, unfair algorithmic behavior, and even discrimination in a legal sense. Some cases, such as salary equity in regards to gender (gender pay gap), stereotypical job perceptions along gendered lines, or biases toward other subgroups sharing specific characteristics in candidate recommenders, can have profound ethical and legal implications. In this survey, we discuss the current state of fairness research considering the fairness definitions (e.g., demographic parity and equal opportunity) used in recruitment-related RSs (RRSs). We investigate from a technical perspective the approaches to improve fairness, like synthetic data generation, adversarial training, protected subgroup distributional constraints, and post-hoc re-ranking. Thereafter, from a legal perspective, we contrast the fairness definitions and the effects of the aforementioned approaches with existing EU and US law requirements for employment and occupation, and second, we ascertain whether and to what extent EU and US law permits such approaches to improve fairness. We finally discuss the advances that RSs have made in terms of fairness in the recruitment domain, compare them with those made in other domains, and outline existing open challenges.

Diverse Sample Generation: Pushing the Limit of Generative Data-Free Quantization.

Generative data-free quantization emerges as a practical compression approach that quantizes deep neural networks to low bit-width without accessing the real data. This approach generates data utilizing batch normalization (BN) statistics of the full-precision networks to quantize the networks. However, it always faces the serious challenges of accuracy degradation in practice. We first give a theoretical analysis that the diversity of synthetic samples is crucial for the data-free quantization, while in existing approaches, the synthetic data completely constrained by BN statistics experimentally exhibit severe homogenization at distribution and sample levels. This paper presents a generic Diverse Sample Generation (DSG) scheme for the generative data-free quantization, to mitigate detrimental homogenization. We first slack the statistics alignment for features in the BN layer to relax the distribution constraint. Then, we strengthen the loss impact of the specific BN layers for different samples and inhibit the correlation among samples in the generation process, to diversify samples from the statistical and spatial perspectives, respectively. Comprehensive experiments show that for large-scale image classification tasks, our DSG can consistently quantization performance on different neural architectures, especially under ultra-low bit-width. And data diversification caused by our DSG brings a general gain to various quantization-aware training and post-training quantization approaches, demonstrating its generality and effectiveness.

Development of Virtual Reality Applications - Digital Display for MSME “Tenun Ikat Bandar” Products

Covid-19 has dramatically impacted the SME sector since April 2020. The government has taken steps to support SMEs as part of the national economic recovery efforts during the pandemic. However, surveys conducted by institutions like BPS, Bappenas, and the World Bank have shown that SMEs need help with loan repayment and meeting expenses for electricity, gas, and employee salaries. SMEs face the challenges of layoffs, limited access to raw materials and capital, declining customers, and production and distribution constraints. Furthermore, changing consumer behavior, business competition, and activity restrictions require proactive measures from SMEs. To address these challenges, a team has proposed a sustainable solution - virtual outlets. Based on virtual reality technology, these outlets allow SMEs to display and sell their products in a digital environment. Visitors can explore every aspect of the virtual outlets, simulating the experience of visiting physical stores without needing in-person visits. This aligns with the new standard of physical/social distancing. Visitors can conveniently support SMEs through online transactions by leveraging virtual outlets, eliminating the need to visit physical stores.

Time-Scale Separation Control for a Class of Current Distribution Strategy in AMBs-Rotor System With Bounded Bus Voltage

In active magnetic bearings (AMBs) systems, good dynamic characteristics and low power consumption are two important concerns, which are difficult to be obtained simultaneously if considering bounded bus voltage. This paper proposes a universal current distribution constraint format and a related time-scale separation control scheme to achieve improvements in both aspects. Considering the bounded bus voltage and symmetrical configuration in one AMB control axis, a double-input three-order nonlinear model is established to describe the AMBs-rotor system. Subsequently, a constraint format is propounded to cover and improve existing current distribution strategies. The power consumption and voltage response are compared between different instances under this constraint format. Then, the double-input model is converted to a single-input cascaded nonlinear system, and a time-scale separation control scheme is designed accordingly. The scheme contains a nominal controller and a dynamic inversion. The nominal controller determines the dynamic characteristics of the AMBs-rotor system. The dynamic inversion compensates for the nonlinearity of the system based on the characteristics of the amplifier. This scheme can preserve the closed-loop stability by only adjusting the time-scale coefficient. Finally, experimental results demonstrate the advantages of the proposed method in aspects of dynamic performance, voltage response, and power consumption.

Collaborative passenger flow control optimization of urban rail transit network based on sliding window mechanism

In order to enhance the operational performance of urban rail transit networks under overloaded conditions, this study proposes a novel network-level collaborative passenger flow strategy based on a sliding window mechanism. The main idea is to categorize arriving passengers at stations into different groups using variable-length windows and allocate them to specific trains. With the aim of minimizing the total passenger extra waiting time under cost budget, we establish a mixed integer programming model with sliding window constraints, flow distribution constraints, and node flow constraints to build the optimal matching relation between passenger flow demands and transportation resources. Specifically, the model incorporates precise quantity constraints related to transfers, facilitating the accurate quantification of both internal and external demands of network passenger flow. Then we use logical constraint transformation and linearization techniques to enable the model to be solved directly by commercial optimization solvers. For large-scale problems, we adopt a rolling horizon approach to improve computing efficiency. To validate the effectiveness of the sliding-window based passenger control method, we conduct case studies including a small designed example and real-world research on the Beijing Subway network. The experimental results demonstrate that our proposed model assists managers in developing flexible and accurate network-level collaborative passenger flow control schemes.

Experimental optimization of various heat sinks using passive thermal management system

Organic phase change materials are extensively researched for passive cooling of electronic components due to the high heat of fusion, however, owing to the issue of thermal conductivity, it is difficult to improve the thermal performance of electronic components. However, the effective thermal performance of modern electronic devices is becoming popular due to thermal constraints of the circuit's non-uniform temperature distribution and high heating power generation. Thus, nanomaterials incorporated into phase change materials (PCMs) to improve thermal conductivity, which aids in heat removal and sustains significant heat sink operational performance for extended periods of time. In current research work, at heating powers of (10–30 W), the thermal performance outcome of three heat sink -configurations such as unfinned heat sink, circular pin-finned heat sink and metallic foam integrated heat sink were investigated with several alumina nanomaterials mass concentrations (0.15, 0.20 and 0.25 wt%) incorporated in phase change materials (for example RT-70HC). All three heat sinks revealed lower base temperature with the addition of alumina NePCM (αRT-70HC) phase change materials in their internal cavity compared to the empty unfinned heat sink. The findings showed good performance of metallic foam integrated heat sink in lowering the temperature & increasing safe functional time at two distinct temperatures. The largest decrease in temperature was found to be 35.76% and the largest growth in maximum functional time was 400% for metallic foam integrated heat sink. Therefore, using alumina nanomaterials in phase change material is recommended to optimize the thermal performance of the passive cooling techniques.

Harmonic Injection Strategy Considering Multiplane Iron Loss Impact With Optimal Magnetizing Flux Distribution for Multiphase Induction Machines

Lagging orientation error detuning induced by iron losses for multi-planes of multiphase IM under non-sinusoidal supply control triggers error in determination of calculated harmonic, fundamental air-gap flux advance angles and harmonic slip frequency. This in-turn leads to inaccurate harmonic injected currents, which are dependent on harmonic multiple of frequency for the corresponding plane with respect to fundamental plane frequency. This paper proposes the novel harmonic injection strategy considering the detuned impact of iron losses by keeping in view the complete iron loss modeling for both harmonic magnetizing flux orientation control (MFOC) and fundamental rotor flux orientation control (RFOC) planes. Furthermore, satisfaction of complete optimal flux distribution constraints for multiphase IM are taken into account which results in preserving optimal flux patterns throughout the mechanical loading particularly at low and medium load conditions. Contrary to other traditional orientation schemes, this paper adopts the MFOC and RFOC for harmonic and fundamental plane respectively for acquiring the aligned and synchronized control of harmonic and fundamental air-gap flux distribution. Detailed experimental results incorporating seven phase induction machine as prototype are utilized to evaluate the harmonic injected currents, advance angles, slip frequencies and induced EMF waveforms which validate the effectiveness of proposed scheme.

A neural network-based distributional constraint learning methodology for mixed-integer stochastic optimization

The use of machine-learning methods helps to improve decision-making in different fields. In particular, the idea of bridging predictions (predictive models) and prescriptions (optimization problems) is gaining attention within the scientific community. One of the main ideas to address this trade-off is the Constraint Learning (CL) methodology, where the structure of the machine learning model can be treated as a set of constraints to be embedded within the optimization problem, establishing the relationship between a direct decision variable x and a response variable y. However, most CL approaches have focused on making point predictions, not considering the statistical and external uncertainty faced in the modeling process. In this paper, we extend the CL methodology to deal with uncertainty in the response variable y. The novel Distributional Constraint Learning (DCL) methodology makes use of a piece-wise linearizable neural network-based model to estimate the parameters of the conditional distribution of y (dependent on decisions x and contextual information), which can be embedded within mixed-integer optimization problems. In particular, we formulate a stochastic optimization problem by sampling random values from the estimated distribution by using a linear set of constraints. In this sense, DCL combines both the predictive performance of the neural network method and the possibility of generating scenarios to account for uncertainty within a tractable optimization model. The behavior of the proposed methodology is tested in the context of electricity systems, where a Virtual Power Plant seeks to optimize its operation, subject to different forms of uncertainty, and with price-responsive consumers.

B1 mapping using pre-learned subspaces for quantitative brain imaging.

To develop a machine learning-based method for estimation of both transmitter and receiver B1 fields desired for correction of the B1 inhomogeneity effects in quantitative brain imaging. A subspace model-based machine learning method was proposed for estimation of B1t and B1r fields. Probabilistic subspace models were used to capture scan-dependent variations in the B1 fields; the subspace basis and coefficient distributions were learned from pre-scanned training data. Estimation of the B1 fields for new experimental data was achieved by solving a linear optimization problem with prior distribution constraints. We evaluated the performance of the proposed method for B1 inhomogeneity correction in quantitative brain imaging scenarios, including T1 and proton density (PD) mapping from variable-flip-angle spoiled gradient-echo (SPGR) data as well as neurometabolic mapping from MRSI data, using phantom, healthy subject and brain tumor patient data. In both phantom and healthy subject data, the proposed method produced high-quality B1 maps. B1 correction on SPGR data using the estimated B1 maps produced significantly improved T1 and PD maps. In brain tumor patients, the proposed method produced more accurate and robust B1 estimation and correction results than conventional methods. The B1 maps were also applied to MRSI data from tumor patients and produced improved neurometabolite maps, with better separation between pathological and normal tissues. This work presents a novel method to estimate B1 variations using probabilistic subspace models and machine learning. The proposed method may make correction of B1 inhomogeneity effects more robust in practical applications.

Feasibility and acceptability of a novel intervention to improve hand hygiene behavior in rural Liberian health facilities

Abstract Hand hygiene is central to the prevention of healthcare-associated infection. In low-income settings, barriers to health worker hand hygiene may include inconsistent availability of hand hygiene supplies at the point of care. However, there is a lack of knowledge of interventions to improve and sustain health worker hand hygiene in these settings. This pilot study evaluates acceptability and feasibility of a personally-worn hand hygiene holster device for improving point-of-care access to alcohol-based handrub (ABHR). Holsters were distributed to clinical staff at a hospital in Liberia in July–September 2021. Data collection included 2,066 structured observations of hand hygiene behavior, six spot checks of supply availability, and focus group discussions with 13 clinical staff. The Integrated Behavioral Model for Water, Sanitation, and Hygiene (IBM-WASH) provided a framework for study design and qualitative analysis. Acceptability of the intervention was high, with users reporting that holsters were comfortable, easy to use, and aligned with their professional identities. Feasibility depended on consistent ABHR availability, which may diminish sustainability of this intervention. The hand hygiene holster is a promising tool for improving health worker hand hygiene behavior, but solutions to ABHR supply chain and distribution constraints are necessary to support sustainability of this intervention.

Data For: Intermittent search, not strict Lévy flight, evolves under relaxed foraging distribution constraints

Data For: Intermittent search, not strict Lévy flight, evolves under relaxed foraging distribution constraints

Semantic Plane-Structure based motion detection with a nonstationary camera

This paper presents a novel method to accurately detect moving objects from a video sequence captured using a nonstationary camera. Although common methods provide effective motion detection for static backgrounds or through only planar-perspective transformation, many detection errors occur when the background contains complex dynamic interferences or the camera undergoes unknown motions. To solve this problem, this study proposed a motion detection method that incorporates temporal motion and spatial structure. In the proposed method, first, spatial semantic planes are segmented, and image registration based on stable background planes is applied to overcome the interferences of the foreground and dynamic background. Thus, the estimated dense temporal motion ensures that small moving objects are not missed. Second, motion pixels are mapped on semantic planes, and then, the spatial distribution constraints of motion pixels, regional shapes and plane semantics, which are integrated into a planar structure, are used to minimise false positives. Finally, based on the dense temporal motion and spatial structure, moving objects are accurately detected. The experimental results on CDnet dataset, Pbi dataset, Aeroscapes dataset, and other challenging self-captured videos under difficult conditions, such as fast camera movement, large zoom variation, video jitters, and dynamic background, revealed that the proposed method can remove background movements, dynamic interferences, and marginal noises and can effectively obtain complete moving objects.© 2017 ElsevierInc.Allrightsreserved.

Research on Demand-Based Scheduling Scheme of Urban Low-Altitude Logistics UAVs

Aiming at the problem of the scheduling scheme of urban low-altitude logistics unmanned aerial vehicles (UAVs), this paper establishes a demand-based UAV scheduling scheme model using an improved simulated annealing algorithm, taking minimizing the cost of distribution as the objective function and considering restrictions such as UAV performance constraints, airspace constraints, and distribution constraints, among others. For verification, actual express data and airspace constraints in Shanghai are taken as examples. Two urban air traffic networks are constructed using road and building data. The analysis results show that the planning scheme of this model is superior to other forecasting models in terms of delivery cost and delivery time. In addition, this model can flexibly calculate the optimal scheduling scheme under the constraints of multiple parameters, according to the requirements of delivery volume, delivery distance, UAV performance, etc.

Spatial Analysis of E-Warong Distribution for Basic Food Social Assistance Program in Surakarta City

The social assistance for the basic food program launched by the Indonesian Ministry of Social Affairs in 2020 seeks social protection for Family Beneficiaries (KPM/Keluarga Peneri-ma Manfaat). One of the distribution constraints is the absence of a database and spatial analysis for the distribution process of social assistance for basic food programs. This study aimed to inventory data on the location of e-warong, visualize, and analyze distance coverage. The method included a survey of the coordinates of the e-Warong location and an analysis of secondary data from the Surakarta City Social Service (Dinas Sosial Kota Surakarta) for March 2022. The data was processed using proximity analysis on Geographic Information System (GIS)-based software to map the affordability of e- Warong services. The results obtained are the creation of a database of e-Warong locations, mapping of accessibility, and the range of e-Warong services to KPM. The accessibility map of the e-warong location is 100 m for the near zone (most covered), 300 m for the medium zone (covered), and 500 m for the far zone (poorly covered). In conclusion, the e-warong service area is partially centered between Pasar Kliwon and Banjarsari. Meanwhile, several locations in Jebres, Serengan, and Laweyan are still unreachable.

Indirect Human Influences in Fear Landscapes: Varying Effects of Moonlight on Small Mammal Activity along Man-Made Gradients of Vegetation Structure

Risk of predation is one of the main constraints of small mammal distribution and foraging activity. Aside from numerical effects on population size due to the presence and abundance of predators, indirect cues, such as vegetation structure and moonlight, determine patterns of activity and microhabitat use by small mammals. Indirect cues are expected to interact, as shading provided by vegetation can suppress the effects of changing moonlight. We analyzed the effects of moonlight levels on the activity patterns of three common small mammal species in Mediterranean habitats, and tested whether moonlight effects were modulated by shadowing associated with the development of tall vegetation due to spontaneous afforestation following land abandonment. A. sylvaticus, a strictly nocturnal species, decreased activity under moonlight with no interactive effects of vegetation cover. C. russula showed no activity change with moonlight levels and M. spretus increased activity, although activity in both species was mostly determined by vegetation cover, that favored it. The effects of moonlight on small mammal activity were not homogeneous among species, nor were the interactive effects of man-made gradients of habitat structure, a fact that will produce community changes along vegetation gradients mediated by varying fear landscapes.

Interior Individual Trajectory Simulation with Population Distribution Constraint

Individual trajectory generation plays an important role in simulation tasks, reconstructing fine-grained mobility behaviors that can be used to evaluate epidemic risks, congestion risks, or commercial profit. Previous research works adopt the Newton’s mechanic-based particle model as their core algorithm, such as the Social Force model. However, real-world human mobility behaviors hardly follow the particle models, especially in the interior scenes where interactions between pedestrians and environments matter. In this article, we propose a Social Force-based trajectory simulator for interior scenarios that improve both trajectory quality and generation speed for interior scenarios. First, we introduce prior scene knowledge to guide the generation process, where pedestrians are armed with exploration behaviors that follow the group-level distribution. It provides more flexibility to simulate complicated human behaviors rather than straight-line movements, generating high-quality individual trajectories. Experiments show that the correlation between the aggregated population distribution of generated trajectories and ground-truth distribution is improved by 11.84% by our method. Second, we optimize the algorithm procedure by introducing a caching mechanism for tenderized intermediate values, along with graph-processing-unit-based implementation. Compared with the baseline Social Force model, we reduced the time consumption by 95%. More importantly, based on our simulation paradigm, we quantitatively evaluate several common mobility interventions in our simulation scenario, which can shed light on better policy designs in public spaces.

Adaptive constraint satisfaction for Markov decision process congestion games: Application to transportation networks

Under the Markov decision process (MDP) congestion game framework, we study the problem of enforcing population distribution constraints on a population of players with stochastic dynamics and coupled congestion costs. Existing research demonstrates that the constraints on the players’ population distribution can be satisfied by enforcing tolls. However, computing the minimum toll value for constraint satisfaction requires accurate modeling of the player’s congestion costs. Motivated by settings where an accurate congestion cost model may be unavailable (e.g. transportation networks), we consider an MDP congestion game with unknown congestion costs. We assume that a constraint-enforcing authority can repeatedly enforce tolls on a population of players that converges to an ϵ-optimal population distribution for any given toll. We then construct a myopic update algorithm to compute the minimum toll value while ensuring that the constraints are satisfied on average. We analyze how the players’ sub-optimal responses to tolls impact the rates of convergence towards the minimum toll value and constraint satisfaction. Finally, we construct a congestion game model for Uber drivers in Manhattan, New York City (NYC) using data from the Taxi and Limousine Commission (TLC) to illustrate how to efficiently reduce congestion while minimizing the impact on driver earnings.

Unsupervised learning-based framework for indirect structural health monitoring using adversarial autoencoder

This paper studies the problem of bridge health monitoring in an unsupervised manner utilizing only the measured responses from a vehicle passing over a bridge. A half-car model along with a simply supported beam is adopted for numerically simulating vehicle–bridge interaction (VBI). Multiple bridge states including healthy bridge and damaged bridge with varying severity at different locations are considered. A data-driven approach based on adversarial autoencoder (AAE) by incorporating the generative capabilities of adversarial autoencoder and pre-processing techniques including frequency filtering and signal averaging is proposed. Vehicle acceleration responses associated only with the healthy bridge state are used for model training. Reconstruction error estimated by the proposed model is adopted as a damage detection index. Along with the detection of the damage in the bridge, the proposed framework is also able to estimate the severity of the damage. The proposed framework also overcomes the limitations of other unsupervised learning approaches such as principal component analysis and autoencoders due to its better representation of data in the latent sub-space with an additional prior distribution constraint. Further, the proposed framework is validated through experimentally obtained data from a laboratory scale bridge model. The contribution of this work is three-fold: First, an adversarial autoencoder-based unsupervised learning framework supplemented by appropriate pre-processing techniques is proposed for drive-by bridge monitoring for the first time, and its implementation is extensively investigated. Second, the superior performance of the proposed AAE framework compared to the competing techniques is demonstrated. Finally, this paper presents one of the early successful attempts of drive-by bridge inspection for monitoring the progressive change in the structure of a bridge. Research presented in this work can potentially open up new opportunities for condition monitoring of bridge networks.

Design of downhill regenerative braking control method for hybrid electric vehicle

In order to improve the downhill regenerative braking control effect of hybrid electric vehicle (HEV), this paper designs a downhill regenerative braking control method for hybrid electric vehicle (HEV). Firstly, the structures of regenerative braking system, auxiliary braking system and eddy current retarder (ECR) are analyzed. Secondly, the braking torque of ECR when HEV downhill braking is calculated. Then, the braking process of HEV is dynamically analyzed, and the braking force distribution constraint of HEV is realized according to the relevant laws and regulations. Finally, the downhill regenerative braking control of HEV is realized based on the braking force distribution coefficient. The results show that when the vehicle speed is up to 120 km·h, the speed control can still be completed within 2 s by using the proposed method, indicating a good regenerative braking effect.

Dual-space configuration synthesis method for rigid-flexible coupled cable-driven parallel mechanisms

As wave motion compensation devices, rigid-flexible coupled cable-driven parallel mechanisms composed of cables and rigid rods can ensure the safety of cargo transportation. Therefore, a dual-space configuration synthesis method for rigid-flexible coupled cable-driven parallel mechanisms is proposed to synthesise configurations with high load-carrying and anti-sway functions. Anti-sway substructures are proposed to reveal the essence of rigid-flexible coupled structures with anti-sway functions. An equivalent rigid-body method and branch distribution constraints are presented to construct the anti-sway substructures and offer iterative conditions for the dual-space configuration synthesis method. Then, thirty-two types of configurations for rigid-flexible coupled cable-driven parallel mechanisms are synthesised through the dual-space configuration synthesis method. Finally, compensation performance evaluation and experimental results for wave motion compensation devices verify the feasibility of the dual-space configuration synthesis method.