A5032248545- Explainable Artificial Intelligence (XAI)
- HIV, Drug Use, Sexual Risk
- Mental Health and Patient Involvement
- Species Distribution and Climate Change
- Gut microbiota and health
- Ecology and Vegetation Dynamics Studies
- Clinical Reasoning and Diagnostic Skills
- Animal Ecology and Behavior Studies
- Model Reduction and Neural Networks
- Plant and animal studies
- Adversarial Robustness in Machine Learning
- Statistical Methods and Inference
- Artificial Intelligence in Healthcare and Education
- Advanced Statistical Process Monitoring
University of Maryland, Baltimore
2024
St. Jude Children's Research Hospital
2024
Johns Hopkins University
2023-2024
University of Baltimore
2023
Yale University
2020
Fairview School District
2019
When humans assemble into face-to-face social networks, they create an extended environment that permits exposure to the microbiome of others, thereby shaping composition and diversity at individual population levels1–6. Here we use comprehensive network mapping detailed sequencing data in 1,787 adults within 18 isolated villages Honduras7 investigate relationship between structure gut composition. Using both species-level strain-level data, show microbial sharing occurs many types, notably...
Abstract When humans assemble into face-to-face social networks, they create an extended environment that permits exposure to the microbiome of other members a population. Social network interactions may thereby also shape composition and diversity at individual population levels. Here, we use comprehensive detailed sequencing data in 1,098 adults across 9 isolated villages Honduras investigate relationship between structure composition. Using both species-level strain-level data, show...
In this multisite prospective study of simulated artificial intelligence (AI)–assisted chest radiograph diagnosis involving 220 physicians, AI explanation type (local vs global) differentially impacted physician diagnostic performance and trust in advice.
Abstract Understanding how abiotic conditions influence dispersal patterns of organisms is important for understanding the degree to which species can track and persist in face changing climate. The goal this study was understand weather pattern multiple nonmigratory grasshopper from lower elevation grassland habitats they complete their life‐cycles higher elevations that extend beyond range limits. Using over a decade weekly spring late‐summer field survey data along an elevational...
As machine learning (ML) gains widespread adoption, practitioners are increasingly seeking means to quantify and control the risk these systems incur. This challenge is especially salient when ML have autonomy collect their own data, such as in black-box optimization active learning, where actions induce sequential feedback-loop shifts data distribution. Conformal prediction has emerged a promising approach uncertainty quantification, but existing variants either fail accommodate sequences...
Students learning the skills of science benefit from opportunities to move between scientific problems and questions they confront mathematical tools available answer solve problems. Indeed, students learn best when are actively engaged in pursuing answers authentic relevant questions. We present an activity teachers can use classroom introduce concepts species richness diversity. break down history logic behind two primary statistical ecologists quantify diversity: Simpson's Shannon's...
We propose \textbf{JAWS}, a series of wrapper methods for distribution-free uncertainty quantification tasks under covariate shift, centered on the core method \textbf{JAW}, \textbf{JA}ckknife+ \textbf{W}eighted with data-dependent likelihood-ratio weights. JAWS also includes computationally efficient \textbf{A}pproximations JAW using higher-order influence functions: \textbf{JAWA}. Theoretically, we show that relaxes jackknife+'s assumption data exchangeability to achieve same finite-sample...