Jump to: Selected Research | Software | Teaching and Workshops | Contact
Hi, I'm Chris!
I am a Ph.D. candidate in Quantitative Psychology at the University of North Carolina at Chapel Hill advised by Daniel Bauer.
Starting in Fall 2023, I will be joining the Department of Psychology at the University of Rhode Island as an Assistant Professor of Quantitative Psychology.
My research is broadly focused on building probabilistic models to understand and predict complex behavioral processes. Toward this end, I develop and disseminate probabilistic machine learning methods that can be used to extract insights from various types of behavioral data including:
As a Ph.D. student, I was supported for three years by a National Science Foundation Graduate Research Fellowship. Before that, I built machine learning models to detect at-risk undergraduate students for the Finish Line Project, an inter-departmental initiative to improve retention of first-generation college students at UNC.
For a complete academic bio, please see my curriculum vitae.
A Deep Learning Algorithm for High-Dimensional Exploratory Item Factor Analysis
We investigate a computationally efficient deep learning algorithm to fit exploratory item response theory (IRT) models (i.e., latent variable models for categorical observed data) when the number of respondents, items, and latent variables are all large. The algorithm performs comparably to and is faster than state-of-the-art IRT estimation methods. Christopher Urban, Daniel BauerPsychometrika, 2021. paper | preprint | code | bibtex |
|
Deep Learning: A Primer for Psychologists
Deep learning has been successfully used to solve complex problems in computer vision and in natural language processing but is rarely used in psychology. In this primer, we provide an overview of deep learning in an effort to bring the benefits of deep learning to psychologists. We use toy examples with R code to demonstrate how foundational deep learning models may be applied to predict important outcomes using the kinds of data sets typically collected by psychologists. Christopher Urban, Kathleen GatesPsychological Methods, 2021. paper | preprint | bibtex |
|
Machine Learning-Based Estimation and Goodness-of-Fit for Large-Scale Confirmatory Item Factor Analysis
I extend my previous work on deep learning-based estimation for exploratory IRT models to the confirmatory setting. I also propose novel computationally efficient methods for assessing model-data fit. Christopher Urban,Master's Thesis, 2021. paper | code | bibtex |
|
DeepIRTools: Deep Learning-Based Estimation and Inference for Large-Scale Item Response Theory
DeepIRTools is a small Pytorch-based Python package that uses scalable deep learning methods to fit a number of different confirmatory and exploratory latent factor models, with a particular focus on item response theory (IRT) models. Graphics processing unit (GPU) support is available for most computations. Christopher Urban, Shara HePython Package, 2022. code | docs | bibtex |
|
E-mail: |