Search

The United States has introduced a special humanitarian parole process for Ukrainian citizens in response to Russia 2022 invasion of Ukraine. To qualify for parole, Ukrainian applicants must have a sponsor in the United States. In collaboration with HIAS, a refugee resettlement agency involved in the parole process, we deployed RUTH (Refugees Uniting Through HIAS), a novel algorithmic matching system that is driven by the relocation preferences of refugees and the priorities of US sponsors. RUTH adapts Thakral Multiple-Waitlist Procedure (MWP) that combines the main FIFO queue with location-specific FIFO queues to effectively manage the preferences of refugees and the supply of community sponsors. RUTH also incorporates various feasibility considerations, such as community capacity religious, and medical needs. The adapted mechanism is envy-free, efficient, and strategy-proof for refugees. Our analysis shows diverse refugee location preferences, not fully explained by observables, highlighting the challenge of predicting preferences directly from them. We use our data for two counterfactual simulations. First, we consider the effects of increased waiting times for refugees on the quality of their matches. We find that with a periodic Top Trading Cycles algorithm, increasing period length from 24 days to 80 days improves the average rank of a refugees match from 3.20 to 2.44. Second, we estimate the arrival rates of sponsors in each location that would be consistent with a long-run steady state. We find that more desirable locations (regarding refugee preferences) require the highest arrival rates, suggesting that preferences might be a useful indicator for investments in sponsorship capacity. Our study highlights the potential for preference-based algorithms such as RUTH to improve the efficiency and fairness of other rapidly deployed humanitarian parole processes., Comment: In Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO23), October 30-November 1, 2023, Boston, MA, USA. 34 pages, including an appendix

Decision making in the public sector centers on delivering resources and services for the common good, emphasizing an expansive set of objectives such as equity and efficiency, beyond immediate short term returns to reflect the broader cares of society and public beneficiaries. Cost-benefit analysis is a prevailing decision-making framework in the public sector that often uses the benefit to cost ratio (BCR) to compare viable alternatives, yet no systematic framework exists for evaluating many alternatives beyond the status quo of doing nothing. We propose a new framework to maximize the BCR for public sector decisions, seeking the largest improvement per marginal deployment of capacity. Requiring a status quo representable through (constrained) decision variables, the framework is generally applicable and useful to a broad set of decision contexts that involve maximizing the BCR for marginal deployments of resources. We demonstrate the applicability of our framework on a compelling case study for the New York City runaway and homeless youth shelter system, an area of high societal need. We represent this problem as a mixed integer linear fractional program (MILFP) and employ Dinkelbach's algorithm that converts the MILFP to a series of linearized mixed-integer optimization problems, making our approach tractable for fairly large problem instances. Our optimization-based algorithmic framework yields data-informed recommendations for making New York City shelter expansion decisions to better serve runaway and homeless youth, and generalizes to reveal managerial insights for optimizing the BCR. More broadly, our algorithmic decision making framework allows for iteration and comparison across multiple potential constraints ensuring action away from the status quo, thereby empowering effective assessment of marginal deployment of additional resources.

Lyman-alpha emitters (LAEs) are excellent probes of the reionization process, as they must be surrounded by large ionized bubbles in order to be visible during the reionization era. Large ionized regions are thought to correspond to over-dense regions and may be protoclusters, making them interesting test-beds for early massive structures. Close associations containing several LAEs are often assumed to mark over-dense, ionized bubbles. Here, we develop the first framework to quantify the ionization and density fields of high-z galaxy associations. We explore the interplay between (i) the large-scale density of a survey field, (ii) Poisson noise due to the small number density of bright sources at high redshifts (z~7), and (iii) the effects of the ionized fraction on the observation of LAEs. We use Bayesian statistics, a simple model of reionization, and a Monte-Carlo simulation to construct a more comprehensive method for calculating the large-scale density of LAE regions than previous works. We find that Poisson noise has a strong effect on the inferred density of a region and show how the ionized fraction can be inferred. We then apply our framework to the strongest association yet identified: Hu et al. (2021) found 14 LAEs in a volume of ~50,000 cMpc^3 inside the COSMOS field at z~7. We show that this is most likely a 2.5-sigma over-density inside of an ionized or nearly ionized bubble. We also show that this LAE association implies that the global ionized fraction is Q = 0.60 (+0.08,-0.09), within the context of a simple reionization model., Comment: 14 pages, 14 figures, submitted to MNRAS

One of the key processes driving galaxy evolution during the Cosmic Dawn is supernova feedback. This likely helps regulate star formation inside of galaxies, but it can also drive winds that influence the large-scale intergalactic medium. Here, we present a simple semi-analytic model of supernova-driven galactic winds and explore the contributions of different phases of galaxy evolution to cosmic metal enrichment in the high-redshift (z > 6) Universe. We show that models calibrated to the observed galaxy luminosity function at z~6-8 have filling factors ~1% at z~6 and ~0.1% at z~12, with different star formation prescriptions providing about an order of magnitude uncertainty. Despite the small fraction of space filled by winds, these scenarios predict an upper limit to the abundance of metal-line absorbers in quasar spectra at z>5 which is comfortably above that currently observed. We also consider enrichment through winds driven by Pop III star formation in minihalos. We find that these can dominate the total filling factor at z>10 and even compete with winds from normal galaxies at z~6, at least in terms of the total enriched volume. But these regions have much lower overall metallicities, because each one is generated by a small burst of star formation. Finally, we show that Compton cooling of these supernova-driven winds at z>6 has only a small effect on the cosmic microwave background., Comment: 13 pages, 8 figures, accepted by MNRAS

One of the most exciting advances of the current generation of telescopes has been the detection of galaxies during the epoch of reionization, using deep fields that have pushed these instruments to their limits. It is essential to optimize our analyses of these fields in order to extract as much information as possible from them. In particular, standard methods of measuring the galaxy luminosity function discard information on large-scale dark matter density fluctuations, even though this large-scale structure drives galaxy formation and reionization during the Cosmic Dawn. Measuring these densities would provide a bedrock observable, connecting galaxy surveys to theoretical models of the reionization process and structure formation. Here, we use existing Hubble deep field data to simultaneously fit the universal luminosity function and measure large-scale densities for each Hubble deep field at $z =$ 6--8 by directly incorporating priors on the large-scale density field and galaxy bias. Our fit of the universal luminosity function is consistent with previous methods but differs in the details. For the first time, we measure the underlying densities of the survey fields, including the most over/under-dense Hubble fields. We show that the distribution of densities is consistent with current predictions for cosmic variance. This analysis on just 17 fields is a small sample of what will be possible with the James Webb Space Telescope, which will measure hundreds of fields at comparable (or better) depths and at higher redshifts.

The epoch of reionization (EoR) offers a unique window into the dawn of galaxy formation, through which high-redshift galaxies can be studied by observations of both themselves and their impact on the intergalactic medium. Line intensity mapping (LIM) promises to explore cosmic reionization and its driving sources by measuring intensity fluctuations of emission lines tracing the cosmic gas in varying phases. Using LIMFAST, a novel semi-numerical tool designed to self-consistently simulate LIM signals of multiple EoR probes, we investigate how building blocks of galaxy formation and evolution theory, such as feedback-regulated star formation and chemical enrichment, might be studied with multi-tracer LIM during the EoR. On galaxy scales, we show that the star formation law and the feedback associated with star formation can be indicated by both the shape and redshift evolution of LIM power spectra. For a baseline model of metal production that traces star formation, we find that lines highly sensitive to metallicity are generally better probes of galaxy formation models. On larger scales, we demonstrate that inferring ionized bubble sizes from cross-correlations between tracers of ionized and neutral gas requires a detailed understanding of the astrophysics that shape the line luminosity--halo mass relation. Despite various modeling and observational challenges, wide-area, multi-tracer LIM surveys will provide important high-redshift tests for the fundamentals of galaxy formation theory, especially the interplay between star formation and feedback by accessing statistically the entire low-mass population of galaxies as ideal laboratories, complementary to upcoming surveys of individual sources by new-generation telescopes., Comment: 24 pages, 15 figures; to appear in ApJ

The New York City (NYC) youth shelter system provides housing, counseling, and other support services to runaway and homeless youth and young adults (RHY). These resources reduce RHYs vulnerability to human trafficking, yet most shelters are unable to meet demand. This paper presents a Discrete Event Simulation (DES) model of a crisis-emergency and drop-in center for LGBTQ+ youth in NYC, which aims to analyze the current operations and test potential capacity expansion interventions. The model uses data from publicly available resources and interviews with service providers and key stakeholders. The simulated shelter has 66 crisis-emergency beds, offers five different support services, and serves on average 1,399 LGBTQ+ RHY per year. The capacity expansion interventions examined in this paper are adding crisis-emergency beds and psychiatric therapists. This application of DES serves as a tool to communicate with policymakers, funders, and service providers potentially having a strong humanitarian impact.

While most methods for solving mixed-integer optimization problems compute a single optimal solution, a diverse set of near-optimal solutions can often lead to improved outcomes. We present a new method for finding a set of diverse solutions by emphasizing diversity within the search for near-optimal solutions. Specifically, within a branch-and-bound framework, we investigated parameterized node selection rules that explicitly consider diversity. Our results indicate that our approach significantly increases the diversity of the final solution set. When compared with two existing methods, our method runs with similar runtime as regular node selection methods and gives a diversity improvement between 12% and 190%. In contrast, popular node selection rules, such as best-first search, in some instances performed worse than state-of-the-art methods by more than 35% and gave an improvement of no more than 130%. Further, we find that our method is most effective when diversity in node selection is continuously emphasized after reaching a minimal depth in the tree and when the solution set has grown sufficiently large. Our method can be easily incorporated into integer programming solvers and has the potential to significantly increase the diversity of solution sets., Comment: 23 pages, 12 figures, submitted to INFORMS Journal on Computing

Recent estimates indicate that there are over 1 million runaway and homeless youth and young adults (RHY) in the United States (US). Exposure to trauma, violence, and substance abuse, coupled with a lack of community support services, puts homeless youth at high risk of being exploited and trafficked. Although access to safe housing and supportive services such as physical and mental healthcare is an effective response to youths vulnerability towards being trafficked, the number of youth experiencing homelessness exceeds the capacity of available housing resources in most US communities. We undertake a RHY-informed, systematic, and data driven approach to project the collective capacity required by service providers to adequately meet the needs of homeless youth in New York City, including those most at risk of being trafficked. Our approach involves an integer linear programming model that extends the multiple multidimensional knapsack problem and is informed by partnerships with key stakeholders. The mathematical model allows for time-dependent allocation and capacity expansion, while incorporating stochastic youth arrivals and length of stays, services provided in a periodic fashion, and service delivery time windows. Our RHY and service provider-centered approach is an important step toward meeting the actual, rather than presumed, survival needs of vulnerable youth, particularly those at-risk of being trafficked.

In 2020 tens of millions of migrants worldwide sought safety due to various push-pull factors such as economic, geographic, demographic, and societal challenges. Large movements of human flows toward national borders constitute a complex humanitarian crisis that requires appropriate preparations of border resources. Motivated by prevailing national agenda issues for improved border support, we identify common operational challenges along the border and discuss how they can be addressed using advanced analytics. Multiple stakeholders, corresponding goals, and complex decisions call for advanced optimization techniques to holistically treat such challenges. This chapter highlights how optimization can be employed to improve both the security and humanitarian aspects of border operations in the context of migration flows., Comment: submitted as a book chapter to Encyclopedia of Optimization (Springer)

Decision support tools enable improved decision-making for challenging decision problems by empowering stakeholders to process, analyze, visualize, and otherwise make sense of a variety of key factors. Their intentional design is a critical component of the value they create. All decision-support tools share in common that there is a complex decision problem to be solved for which decision-support is useful, and moreover, that appropriate analytics expertise is available to produce solutions to the problem setting at hand. When well-designed, decision support tools reduce friction and increase efficiency in providing support for the decision-making process, thereby improving the ability of decision-makers to make quality decisions. On the other hand, the presence of overwhelming, superfluous, insufficient, or ill-fitting information and software features can have an adverse effect on the decision-making process and, consequently, outcomes. We advocate for an innovative, and perhaps overlooked, approach to designing effective decision support tools: genuinely listening to the project stakeholders, to ascertain and appreciate their real needs and perspectives. By prioritizing stakeholder needs, a foundation of mutual trust and understanding is established with the design team. We maintain this trust is critical to eventual tool acceptance and adoption, and its absence jeopardizes the future use of the tool, which would leave its analytical insights for naught. We discuss examples across multiple contexts to underscore our collective experience, highlight lessons learned, and present recommended practices to improve the design and eventual adoption of decision dupport tools.

Cosmic variance from large-scale structure will be a major source of uncertainty for galaxy surveys at z > 6, but that same structure will also provide an opportunity to identify and study dense environments in the early Universe. Using a robust model for galaxy clustering, we directly incorporate large-scale densities into an inference framework that simultaneously measures the high-z (z > 6) UV luminosity function and the average matter density of each distinct volume in a survey. Through this framework, we forecast the performance of several major upcoming James Webb Space Telescope (JWST) galaxy surveys. We find that they can constrain field matter densities down to the theoretical limit imposed by Poisson noise and unambiguously identify over-dense (and under-dense) regions on transverse scales of tens of comoving Mpc. We also predict JWST will measure the luminosity function with a precision at z = 12 comparable to existing Hubble Space Telescope's constraints at z = 8 (and even better for the faint-end slope). We also find that wide-field surveys are especially important in distinguishing luminosity function models.

Employment outcomes of resettled refugees depend strongly on where they are placed inside the host country. Each week, a resettlement agency is assigned a batch of refugees by the United States government. The agency must place these refugees in its local affiliates, while respecting the affiliates' yearly capacities. We develop an allocation system that suggests where to place an incoming refugee, in order to improve total employment success. Our algorithm is based on two-stage stochastic programming and achieves over 98 percent of the hindsight-optimal employment, compared to under 90 percent of current greedy-like approaches. This dramatic improvement persists even when we incorporate a vast array of practical features of the refugee resettlement process including indivisible families, batching, and uncertainty with respect to the number of future arrivals. Our algorithm is now part of the Annie MOORE optimization software used by a leading American refugee resettlement agency., Comment: Expanded related work, added experiments with bootstrapped arrivals in Section 7.2, added various experiments in the appendix

Human trafficking is a widespread and compound social, economic, and human rights issue occurring in every region of the world. While there have been an increasing number of anti-human trafficking works from the Operations Research and Analytics domains in recent years, no systematic review of this literature currently exists. We fill this gap by providing a systematic literature review that identifies and classifies the body of Operations Research and Analytics research related to the anti-human trafficking domain, thereby illustrating the collective impact of the field to date. We classify 142 studies to identify current trends in methodologies, theoretical approaches, data sources, trafficking contexts, target regions, victim-survivor demographics, and focus within the well-established 4Ps principles. Using these findings, we discuss the extent to which the current literature aligns with the global demographics of human trafficking and identify existing research gaps to propose an agenda for Operations Research and Analytics researchers., Comment: 34 pages, 10 Figures, 2 Tables

Using a subsample of the Bulge Asymmetries and Dynamical Evolution (BAaDE) survey of stellar SiO masers, we explore the prevalence and characteristics of $^{28}$SiO $J=1-0, v=0$ emission. We identify 90 detections of maser, thermal, or composite $^{28}$SiO $J=1-0, v=0$ emission out of approximately 13,000 candidate spectra from the NSF's Karl G. Jansky Very Large Array (VLA). We find that the detected sources are likely asymptotic giant branch (AGB) stars belonging to a bright, foreground Milky Way stellar disk population. For the 32 sources showing thermal components, we extract values for outflow velocity by fitting thermal line profiles. We find a range of circumstellar envelope expansion velocities, and compare to previously recorded OH and CO expansion velocities. This preliminary survey is already the largest study of stellar ground-vibrational-state SiO masers to date, and will be expanded to include the entire VLA BAaDE dataset when data reduction for the 18,988 target sources is completed., Comment: 23 pages, 8 figures

Transit monitoring is a preventative approach used to identify possible cases of human trafficking prior to exploitation while an individual is in transit or before one crosses a border. Transit monitoring is often conducted by non-governmental organizations (NGOs) who train staff to identify and intercept suspicious activity. Love Justice International (LJI) is a well-established NGO that has been conducting transit monitoring for years along the Nepal-India border at multiple monitoring stations. In partnership with LJI, we developed a system that uses data envelopment analysis (DEA) to help LJI decision-makers evaluate the performance of these stations at intercepting potential human-trafficking victims given the amount of resources (e.g. staff, etc.) available and make specific operational improvement recommendations. Our model consists of 91 decision-making units (DMUs) from 7 stations over 13 quarters and considers three inputs, four outputs, and 3 homogeneity criteria. Using this model we identified efficient stations, compared rankings of station performance, and recommended strategies to improve efficiency. To the best of our knowledge, this is the first application of DEA in the anti-human trafficking domain.

Cosmic variance is the intrinsic scatter in the number density of galaxies due to fluctuations in the large-scale dark matter density field. In this work, we present a simple analytic model of cosmic variance in the high redshift Universe ($z\sim5-15$). We assume that galaxies grow according to the evolution of the halo mass function, which we allow to vary with large-scale environment. Our model produces a reasonable match to the observed ultraviolet luminosity functions in this era by regulating star formation through stellar feedback and assuming that the UV luminosity function is dominated by recent star formation. We find that cosmic variance in the UVLF is dominated by the variance in the underlying dark matter halo population, and not by differences in halo accretion or the specifics of our stellar feedback model. We also find that cosmic variance dominates over Poisson noise for future high-$z$ surveys except for the brightest sources or at very high redshifts ($z \gtrsim 12$). We provide a linear approximation of cosmic variance for a variety of redshifts, magnitudes, and survey areas through the public Python package galcv. Finally, we introduce a new method for incorporating priors on cosmic variance into estimates of the galaxy luminosity function and demonstrate that it significantly improves constraints on that important observable., Comment: 16 pages, 15 figures, accepted by MNRAS

Human trafficking, the commercial exploitation of individuals, is a gross violation of human rights and harms societies, economies, health and development. The related disciplines of Operations Management (OM), Analytics, and Operations Research (OR) are uniquely positioned to support trafficking prevention and intervention efforts by efficiently evaluating a plethora of decision alternatives, and providing quantitative, actionable insights. As operations and analytical efforts in the counter-trafficking field emerge, it is imperative to grasp subtle yet distinctive nuances associated with human trafficking. This note is intended to inform those practitioners working in the field by highlighting key features of human trafficking activity. We grouped nine themes around three broad categories: (1) representation of human trafficking, (2) consideration of survivors and communities, and (3) analytics related. These insights are derived from our collective experience in working in this area and substantiated by domain expertise.

We present new sub-arcsecond-resolution Karl G. Jansky Very Large Array (VLA) imaging at 10 GHz of 155 ultra-luminous ($L_{\rm bol}\sim10^{11.7-14.2} L_\odot$) and heavily obscured quasars with redshifts $z \sim0.4-3$. The sample was selected to have extremely red mid-infrared (MIR)-optical color ratios based on data from Wide-Field Infrared Survey Explorer (WISE) along with a detection of bright, unresolved radio emission from the NRAO VLA Sky Survey (NVSS) or Faint Images of the Radio Sky at Twenty-Centimeters (FIRST) Survey. Our high-resolution VLA observations have revealed that the majority of the sources in our sample (93 out of 155) are compact on angular scales $<0.2^{\prime \prime}$ ($\leq 1.7$ kpc at $z \sim2$). The radio luminosities, linear extents, and lobe pressures of our sources are similar to young radio active galactic nuclei (AGN; e.g., Gigahertz Peaked Spectrum, GPS, and Compact Steep Spectrum, CSS, sources), but their space density is considerably lower. Application of a simple adiabatic lobe expansion model suggests relatively young dynamical ages ($\sim10^{4-7}$ years), relatively high ambient ISM densities ($\sim1-10^4$ cm$^{-3}$), and modest lobe expansion speeds ($\sim30-10,000$ km s$^{-1}$). Thus, we find our sources to be consistent with a population of newly triggered, young jets caught in a unique evolutionary stage in which they still reside within the dense gas reservoirs of their hosts. Based on their radio luminosity function and dynamical ages, we estimate only $\sim20\%$ of classical large scale FRI/II radio galaxies could have evolved directly from these objects. We speculate that the WISE-NVSS sources might first become GPS or CSS sources, of which some might ultimately evolve into larger radio galaxies., Comment: Accepted for publication in ApJ, 38 pages, 14 figures, 6 tables

We present a global optimization approach for solving the maximum a-posteriori (MAP) clustering problem under the Gaussian mixture model.Our approach can accommodate side constraints and it preserves the combinatorial structure of the MAP clustering problem by formulating it asa mixed-integer nonlinear optimization problem (MINLP). We approximate the MINLP through a mixed-integer quadratic program (MIQP) transformation that improves computational aspects while guaranteeing $\epsilon$-global optimality. An important benefit of our approach is the explicit quantification of the degree of suboptimality, via the optimality gap, en route to finding the globally optimal MAP clustering. Numerical experiments comparing our method to other approaches show that our method finds a better solution than standard clustering methods. Finally, we cluster a real breast cancer gene expression data set incorporating intrinsic subtype information; the induced constraints substantially improve the computational performance and produce more coherent and bio-logically meaningful clusters.

The InfraRed Imaging Spectrograph (IRIS) is a first-light instrument for the Thirty Meter Telescope (TMT). It combines a diffraction limited imager and an integral field spectrograph. This paper focuses on the electrical system of IRIS. With an instrument of the size and complexity of IRIS we face several electrical challenges. Many of the major controllers must be located directly on the cryostat to reduce cable lengths, and others require multiple bulkheads and must pass through a large cable wrap. Cooling and vibration due to the rotation of the instrument are also major challenges. We will present our selection of cables and connectors for both room temperature and cryogenic environments, packaging in the various cabinets and enclosures, and techniques for complex bulkheads including for large detectors at the cryostat wall.

We present the first results from the BAaDE (Bulge Asymmetries and Dynamic Evolution) survey. Though only a subset of the complete survey ($\sim$2700 out of $\sim$20000 final sources), our data comprise the largest radio kinematic survey to date of stellar SiO masers observed toward the Galactic bulge and plane from $-15^\circ < l < +12^\circ$ and $-6^\circ < b < +6^\circ$. Our sources include a substantial number of line-of-sight (LoS) velocities in high extinction regions within $\pm 1^\circ$ of the Galactic plane. When matched with 2MASS photometry, our radio-detected sample lies significantly brighter than and red-ward of the first red giant branch tip, reaching extremes of $(J-Ks)_0 > 8$, colors consistent with Mira variables and mass losing AGB stars. We see a clean division into two kinematic populations: a kinematically cold ($\sigma \sim 50\rm\ km\ sec^{-1}$) population that we propose is in the foreground disk, consisting of giants with 2MASS $Ks<5.5$, and a kinematically hot ($\sigma \sim 100\rm\ km\ sec^{-1}$) candidate bulge/bar population for most giants with $Ks>5.5$. Only the kinematically hot giants with $Ks>5.5$ include the reddest stars. Adopting 8.3 kpc to the Galactic Center, and correcting for foreground extinction, we find that most of the sources have $M_{bol} \sim -5$, consistent with their being luminous, and possibly intermediate age, AGB stars. We note some tension between the possibly intermediate age of the kinematically hot population, and its high velocity dispersion compared to the disk., Comment: 18 pages, 17 Figures; Submitted to ApJ (Accepted May 7)

Variational inference methods for latent variable statistical models have gained popularity because they are relatively fast, can handle large data sets, and have deterministic convergence guarantees. However, in practice it is unclear whether the fixed point identified by the variational inference algorithm is a local or a global optimum. Here, we propose a method for constructing iterative optimization algorithms for variational inference problems that are guaranteed to converge to the $\epsilon$-global variational lower bound on the log-likelihood. We derive inference algorithms for two variational approximations to a standard Bayesian Gaussian mixture model (BGMM). We present a minimal data set for empirically testing convergence and show that a variational inference algorithm frequently converges to a local optimum while our algorithm always converges to the globally optimal variational lower bound. We characterize the loss incurred by choosing a non-optimal variational approximation distribution suggesting that selection of the approximating variational distribution deserves as much attention as the selection of the original statistical model for a given data set., Comment: 27 pages, 3 figures