About

Professor Byung-Eun Kim is a faculty member in the Department of Animal & Avian Sciences at the University of Maryland. His research focuses on elucidating the molecules and mechanisms required for copper homeostasis at cellular, organismal, and behavioral levels. He investigates systemic copper signaling pathways, inter-organ signaling factors related to cardiac copper deficiency, and the role of the Ctr1 copper ion channel as a potential risk factor for cardiomyopathy. Additionally, his work includes understanding molecular mechanisms of copper uptake and distribution in the model organism C. elegans, as well as identifying novel copper-regulated genes and systemic pathways that have similar functions in mammals.

Research topics

• Computer Science
• Artificial Intelligence
• Political Science
• Computer Security
• Machine Learning
• Computer network
• Internet privacy
• Statistics
• Environmental health
• Psychology
• Business
• Public relations
• Finance
• World Wide Web
• Telecommunications
• Medicine

Selected publications

• Judgment of Social Traits and Emotions of “Neutral” Avatars

  Cyberpsychology Behavior and Social Networking · 2025-01-06 · 3 citations

  articleOpen access

  With the rapid advance of technology, human interactions with virtual avatars in simulated social environments are becoming increasingly common. The aim of the current study was to examine users’ perception of social traits and emotions of “neutral,” expressionless avatars using an open-source collection. These avatars represented different ethnicities, genders, and occupations via visual features including skin tone, facial structure, and apparel. We hypothesized that the social evaluation of “neutral” avatars would be influenced by these visual features. In two online studies, we asked survey participants ( N = 225) to identify and rate the social traits and determine the expressed emotion of avatars. Female avatars were rated more attractive, trustworthy, friendly, and less aggressive than male avatars. Black avatars were rated more attractive, friendly, and trustworthy in comparison to White avatars. Avatars in martial uniforms were rated as more aggressive and less friendly than avatars in non-martial uniforms. In turn, non-martial uniformed avatars were rated higher in trustworthiness and intelligence than avatars in martial uniforms and avatars without uniforms. These results suggest that users attribute social traits and emotions to “neutral” avatars. These findings have implications for the design of tasks and products that rely on the selection of avatars in virtual reality.

  PublisherOA PDFDOI

• Large Language Models for Interpretable Mental Health Diagnosis

  ArXiv.org · 2025-01-13 · 1 citations

  preprintOpen access1st authorCorresponding

  We propose a clinical decision support system (CDSS) for mental health diagnosis that combines the strengths of large language models (LLMs) and constraint logic programming (CLP). Having a CDSS is important because of the high complexity of diagnostic manuals used by mental health professionals and the danger of diagnostic errors. Our CDSS is a software tool that uses an LLM to translate diagnostic manuals to a logic program and solves the program using an off-the-shelf CLP engine to query a patient's diagnosis based on the encoded rules and provided data. By giving domain experts the opportunity to inspect the LLM-generated logic program, and making modifications when needed, our CDSS ensures that the diagnosis is not only accurate but also interpretable. We experimentally compare it with two baseline approaches of using LLMs: diagnosing patients using the LLM-only approach, and using the LLM-generated logic program but without expert inspection. The results show that, while LLMs are extremely useful in generating candidate logic programs, these programs still require expert inspection and modification to guarantee faithfulness to the official diagnostic manuals. Additionally, ethical concerns arise from the direct use of patient data in LLMs, underscoring the need for a safer hybrid approach like our proposed method.

  PublisherOA PDFDOI

• FairQuant: Certifying and Quantifying Fairness of Deep Neural Networks

  arXiv (Cornell University) · 2024-09-05 · 1 citations

  preprintOpen access1st authorCorresponding

  We propose a method for formally certifying and quantifying individual fairness of deep neural networks (DNN). Individual fairness guarantees that any two individuals who are identical except for a legally protected attribute (e.g., gender or race) receive the same treatment. While there are existing techniques that provide such a guarantee, they tend to suffer from lack of scalability or accuracy as the size and input dimension of the DNN increase. Our method overcomes this limitation by applying abstraction to a symbolic interval based analysis of the DNN followed by iterative refinement guided by the fairness property. Furthermore, our method lifts the symbolic interval based analysis from conventional qualitative certification to quantitative certification, by computing the percentage of individuals whose classification outputs are provably fair, instead of merely deciding if the DNN is fair. We have implemented our method and evaluated it on deep neural networks trained on four popular fairness research datasets. The experimental results show that our method is not only more accurate than state-of-the-art techniques but also several orders-of-magnitude faster.

  PublisherOA PDFDOI

• Using a Stopping Rule to Optimize Cost-Quality Tradeoffs in a Large, Mixed-Mode Survey: A Simulation Study

  Journal of Official Statistics · 2024-11-11 · 2 citations

  articleOpen access

  Rising nonresponse and the increasing costs of conducting surveys are creating pressure on survey organizations to efficiently allocate resources. The challenge is to produce the highest quality data possible within a fixed budget. We use a stopping rule designed to minimize a function of cost and errors. The rule is based on the product of predicted costs and the predicted mean squared error of a survey estimate. We simulate the impact of implementing the stopping rule on the 2019 US Survey of Doctorate Recipients (SDR), which is a longitudinal survey conducted every two years in the US using web, mail, and CATI. We vary the types of models used to generate the input predictions (parametric regression vs. nonparametric tree models) and the timing of the implementation of the rule. We found that the modeling approach made less difference, while the timing of the implementation of the stopping rule made a large difference in outcomes. The rule is multivariate and optimizes outcomes for two variables. It performed better for one variable, leading to reduced costs and only small increases in errors. The other variable had larger error increases.

  PublisherDOI

• Instructional System Coherence: A Scoping Literature Review

  RAND Corporation eBooks · 2024-01-01 · 1 citations

  bookOpen access

  This report describes the authors' findings from a scoping literature review on the topic of instructional system coherence. The authors reviewed 77 pieces on coherence published from 1990 to 2023, including articles, reports, books, dissertations, and conference papers. They consider overall methods and takeaways from this body of literature.

  PublisherOA PDFDOI

• Hidden population size estimation and diagnostics using two respondent‐driven samples with applications in Armenia

  Biometrical Journal · 2023-03-06 · 3 citations

  articleOpen access1st author

  Estimating the size of hidden populations is essential to understand the magnitude of social and healthcare needs, risk behaviors, and disease burden. However, due to the hidden nature of these populations, they are difficult to survey, and there are no gold standard size estimation methods. Many different methods and variations exist, and diagnostic tools are needed to help researchers assess method-specific assumptions as well as compare between methods. Further, because many necessary mathematical assumptions are unrealistic for real survey implementation, assessment of how robust methods are to deviations from the stated assumptions is essential. We describe diagnostics and assess the performance of a new population size estimation method, capture-recapture with successive sampling population size estimation (CR-SS-PSE), which we apply to data from 3 years of studies from three cities and three hidden populations in Armenia. CR-SS-PSE relies on data from two sequential respondent-driven sampling surveys and extends the successive sampling population size estimation (SS-PSE) framework by using the number of individuals in the overlap between the two surveys and a model for the successive sampling process to estimate population size. We demonstrate that CR-SS-PSE is more robust to violations of successive sampling assumptions than SS-PSE. Further, we compare the CR-SS-PSE estimates to population size estimations using other common methods, including unique object and service multipliers, wisdom of the crowd, and two-source capture-recapture to illustrate volatility across estimation methods.

  PublisherOA PDFDOI

• Building Organizational and Individual Capacities in State and Local Education Agencies: The First Five Years of the Center to Improve Social and Emotional Learning and School Safety

  RAND Corporation eBooks · 2023-01-01

  bookOpen accessSenior author

  In this report, the authors explore how the Center to Improve Social and Emotional Learning and School Safety designed and implemented its technical assistance supports and how these supports contributed to organizational and individual capacity building among technical assistance recipients. The report concludes with implications for technical assistance providers in the development and provision of capacity-building support.

  PublisherOA PDFDOI

• Promoting Publications Through Plastic Surgery Journal Instagram Accounts

  Annals of Plastic Surgery · 2023-03-04 · 3 citations

  article

  PURPOSE: Journals are increasingly using social media to increase article engagement. We aim to determine the impact of Instagram promotion on, and identify social media tools that effectively enhance, plastic surgery article engagement and impact. METHODS: Instagram accounts for Plastic and Reconstructive Surgery , Annals of Plastic Surgery , Aesthetic Surgery Journal , and Aesthetic Plastic Surgery were reviewed for posts published by February 8, 2022. Open access journal articles were excluded. Post caption word count and number of likes, tagged accounts, and hashtags were recorded. Inclusion of videos, article links, or author introductions was noted. All articles from journal issues published between the dates of the first and last posts promoting articles were reviewed. Altmetric data approximated article engagement. Citation numbers from the National Institutes of Health iCite tool approximated impact. Differences in engagement and impact of articles with and without Instagram promotion were compared by Mann-Whitney U tests. Univariate and multivariable regressions identified factors predictive of more engagement (Altmetric Attention Score, ≥5) and citations (≥7). RESULTS: A total of 5037 articles were included, with 675 (13.4%) promoted on Instagram. Of posts featuring articles, 274 (40.6%) included videos, 469 (69.5%) included article links, and 123 included (18.2%) author introductions. Promoted articles had higher median Altmetric Attention Scores and citations ( P < 0.001). On multivariable analysis, using more hashtags predicted higher article Altmetric Attention Scores (odds ratio [OR], 1.85; P = 0.002) and more citations (OR, 1.90; P < 0.001). Including article links (OR, 3.52; P < 0.001) and tagging more accounts (OR, 1.64; P = 0.022) predicted higher Altmetric Attention Scores. Including author introductions negatively predicted Altmetric Attention Scores (OR, 0.46; P < 0.001) and citations (OR, 0.65; P = 0.047). Caption word count had no significant impact on article engagement or impact. CONCLUSIONS: Instagram promotion increases plastic surgery article engagement and impact. Journals should use more hashtags, tag more accounts, and include manuscript links to increase article metrics. We recommend that authors promote on journal social media to maximize article reach, engagement, and citations, which positively impacts research productivity with minimal additional effort in designing Instagram content.

  PublisherDOI

• Adversarial Machine Learning for NextG Covert Communications Using Multiple Antennas

  Entropy · 2022-07-29 · 3 citations

  articleOpen access1st author

  This paper studies the privacy of wireless communications from an eavesdropper that employs a deep learning (DL) classifier to detect transmissions of interest. There exists one transmitter that transmits to its receiver in the presence of an eavesdropper. In the meantime, a cooperative jammer (CJ) with multiple antennas transmits carefully crafted adversarial perturbations over the air to fool the eavesdropper into classifying the received superposition of signals as noise. While generating the adversarial perturbation at the CJ, multiple antennas are utilized to improve the attack performance in terms of fooling the eavesdropper. Two main points are considered while exploiting the multiple antennas at the adversary, namely the power allocation among antennas and the utilization of channel diversity. To limit the impact on the bit error rate (BER) at the receiver, the CJ puts an upper bound on the strength of the perturbation signal. Performance results show that this adversarial perturbation causes the eavesdropper to misclassify the received signals as noise with a high probability while increasing the BER at the legitimate receiver only slightly. Furthermore, the adversarial perturbation is shown to become more effective when multiple antennas are utilized.

  PublisherOA PDFDOI

• The Value of Science: Special Theme

  Harvard Data Science Review · 2022-04-28 · 1 citations

  articleOpen access

  This special report on the Value of Science: Data, Products, and Use reflects the results of a conference intended to advance understanding of the value of data by showcasing new data, products, and use resulting from recent data investments in science policy.

  PublisherOA PDFDOI

Frequent coauthors

• Ellis Meng

  14 shared

• Gary W. Chien

  Kaiser Permanente

  13 shared

• Şennur Ulukuş

  10 shared

• Tugba Erpek

  Virginia Tech

  10 shared

• Yalin E. Sagduyu

  8 shared

• Kemal Davaslıoğlu

  7 shared

• Jonathan T. W. Kuo

  6 shared

• Seth A. Hara

  Mayo Clinic

  6 shared

Education

• PhD, Statistics

  University of California Los Angeles

  2017