Rebecca Nugent is the Stephen E. and Joyce Fienberg Professor of Statistics & Data Science, the Department Head for the Carnegie Mellon Statistics & Data Science Department, and an affiliated faculty member of the Block Center for Technology and Society. She received her PhD in Statistics from the University of Washington in 2006. Prior to that, she received her B.A. in Mathematics, Statistics, and Spanish from Rice University and her M.S. in Statistics from Stanford University. She was won several national and university teaching awards including the American Statistical Association Waller Award for Innovation in Statistics Education and serves as one of the co-editors of the Springer Texts in Statistics. She recently served on the National Academy of Sciences study on Envisioning the Data Science Discipline: The Undergraduate Perspective and is the co-chair of the current NAS study Improving Defense Acquisition Workforce Capability in Data Use. She is the Founding Director of the Statistics & Data Science Corporate Capstone program, an experiential learning initiative that matches groups of faculty and students with data science problems in industry, non-profits, and government organizations. She has worked extensively in clustering and classification methodology with an emphasis on high-dimensional, big data problems and record linkage applications. Her current research focus is the development and deployment of low-barrier data analysis platforms that allow for adaptive instruction and the study of data science as a science.
Eduard Gröller is Professor at the Institute of Visual Computing & Human-Centered Technology (VC&HCT), TU Wien, where he is heading the Research Unit of Computer Graphics. He is a scientific proponent and key researcher of the VRVis research center. The center performs applied research in visualization, rendering, and visual analysis. Dr. Gröller is Adjunct Professor of Computer Science at the University of Bergen, Norway. His research interests include computer graphics, visualization, and visual computing. He became a fellow of the Eurographics Association in 2009. Dr. Gröller is the recipient of the Eurographics 2015 Outstanding Technical Contributions Award and of the IEEE VGTC 2019 Technical Achievement Award.
Certain Uncertainties in Visual Data Science
Uncertainty in visual data science concerns challenges, ambiguities, and potential errors that arise when interpreting, analyzing, or representing data visually. It is an inherent part of the data or models, stemming from various sources like measurement errors, incomplete or noisy data, or limitations in predictive models. Several examples are discussed in more detail. In the first example, Fuzzy Spreadsheets extend traditional spreadsheets by allowing cells to hold value distributions, enabling users to analyze uncertainty and perform what-if scenarios. This approach keeps the familiar interface while adding insights into sensitivity and robustness. The second example from hazard overview and risk assessment discusses an interactive web application to help the public understand personal flood risks more intuitively. Personalized, simplified 3D visualizations show relevant flood hazard information, helping users assess their property’s exposure and vulnerability. The third example deals with documentary-style molecular visualizations that use real-time rendering and structured storytelling to explain complex biological models. This approach combines automated visual tours with text-to-speech narration, making scientific content more accessible to general audiences, where storytelling involves guidance and uncertainty simultaneously. In the last part of the talk possible future research directions on uncertainties in visual data science touching confidence and trust will be elaborated.
Katrijn Van Deun is a Professor in Data Science for the Social and Behavioral sciences at Tilburg University, affiliated to the department of Methodology & Statistics since 2014. She received her PhD in Psychometrics from the KU Leuven in 2005 and worked as a postdoc in computational biology at that same university. Her research focuses on the development of statistical methods for high-dimensional Multiview data, to support the development of personalized multidisciplinary treatment plans in collaboration with behavioral and medical scientists. She is an elected member of the International Statistical Institute, recipient of a prestigious personal Vici grant from the Dutch Research Council, and winner of the Best Teacher award of the joint bachelor Data Science of Tilburg University and Eindhoven University of Technology.
Regularized exploratory approximate factor analysis for easy analysis of complex data
Non-observable constructs such as personality, intelligence, and well-being are at the core of research on human behaviour and cognition. Latent variable methods (e.g., factor analysis, structural equation modelling) are therefore an indispensable tool for research in the social and behavioural sciences. These methods are known to work well when the number of parameters to estimate is relatively small compared to the sample size. However, modern research relies on large collections of data, including multidisciplinary approaches where several blocks of variables have been measured on the same persons. Currently available latent variable methods are restrictive in their use, often not allowing to analyze high-dimensional data and/or taking the multi-block structure into account.
Here, we propose a regularized latent variable method that addresses these issues by relying on an approximate factor analysis approach and a strong computational framework inspired by alternating optimization and the alternating direction method of multipliers. We illustrate the (superior) performance of our method with simulated and real data, including psychological questionnaire and large genetic data obtained from patients with several psychiatric disorders.