Research overview:
My research program contains three main themes:
1. Bayesian models and algorithms for large and dependent data: In modern applications, dependencies in the data are often unavoidable and appear concurrently (e.g., multivariate, time series, functional, and/or spatial data). A focal point of our research is the careful development of Bayesian models and algorithms for reliable and scalable inference with dependent data, including cases with restricted or nonstandard support (e.g., discrete, mixed, or spherical data).
2. Modeling, synthesis, imputation of mixed data: Mixed datasets, including continuous, count, ordinal, and unordered categorical variables, are abundant and challenging to model jointly—especially in the presence of informative missingness. Our research innovations include nonparametric and semiparametric modeling frameworks that provide correct support, substantial joint modeling flexibility with theoretical guarantees, and highly convenient and scalable computing. This work is particularly useful for prediction, data synthesis for privacy protection, and imputation of missing data.
3. Predictive inference for more interpretable uncertainty quantification: Almost universally, statistical measures of uncertainty are linked to unobservable parameters in complex models, which inhibits clear communication to both scientific and public audiences. Our work introduces a unified, posterior predictive framework for versatile, observation-driven uncertainty quantification. This framework is widely applicable to classical statistical problems as well as complex Bayesian models, and includes important special cases such as subset selection, detection of critical windows of susceptibility (e.g., to adverse environmental exposures), and inference with missing data.
This research is motivated by urgent and open questions in public health, epidemiology and environmental justice, physical activity data, economics, and finance.
My research has been recognized with a Young Investigator Award (Army Research Office), the inaugural Blackwell-Rosenbluth award (International Society of Bayesian Analysis) for “junior researchers in different areas of Bayesian statistics”, and several paper and dissertation awards.
Additional background on me from Cornell CALS and Cornell Bowers CIS.
News:
Tenure awarded at Cornell! (11/01/2024)
Monte Carlo inference for semiparametric Bayesian regression accepted to JASA! Joint work with Bohan Wu (08/19/2024)
Software available on CRAN: SeBR
Stories from Phys.org, National Tribune, Cornell Chronicle, Mirage, and Cornell CALS
Spatial variability in relationships between early childhood lead exposure and standardized test scores in 4th grade North Carolina public school students (2013-2016) accepted to Environmental Health Perspectives! Joint work with Mercedes Bravo, Dominique Zephyr, Joe Feldman, Kathy Ensor, and Marie Lynn Miranda (08/13/2024)
Facilitating heterogeneous effect estimation via statistically efficient categorical modifiers submitted! (08/01/2024)
R package lmabc with documentation and examples in the online vignette
Lindley Prize: Honorable Mention for the paper Semiparametric Functional Factor Models with Bayesian Rank Selection with Tony Canale (07/06/2024)
We’re moving! Beginning July 1, 2024, I will be associate professor in the Department of Statistics and Data Science at Cornell University.
Evaluating integration of letter fragments through contrast and spatially targeted masking published in Journal of Vision! Joint work with Sherry Zhang, Jack Morrison, Thomas Sun, and Ernest Greene (06/10/2024)
Gaussian Copula Models for Nonignorable Missing Data Using Auxiliary Marginal Quantiles submitted! Joint work with Joe Feldman and Jerry Reiter (06/05/2024)
Nonparametric Copula Models for Multivariate, Mixed, and Missing Data accepted to JMLR! Joint work with Joe Feldman (05/18/2024)
R code is available on GitHub!
Ultra-efficient MCMC for Bayesian longitudinal functional data analysis accepted to JCGS! Joint work with Thomas Sun (05/16/2024)
R code is available on GitHub!
Promoted to associate professor with tenure at Rice University, effective July 1! (05/10/2024)
Some brief comments on editorial service and the peer-review process (05/09/2024)
New software available: lmabc for linear regression with categorical covariates! Joint work with Prayag Gordy, Virginia Baskin, Jai Uparkar, and Caleb Fikes (04/01/2024)
Congratulations to Thomas Sun on his ASA Statistical Computing and Statistical Graphics Sections student paper award! (01/15/2024)
Congratulations to John Zito on his ASA Section on Business and Economic Statistics student paper award! (01/15/2024)
Regression with race-modifiers: towards equity and interpretability submitted! (12/17/2023)
Semiparametric Functional Factor Models with Bayesian Rank Selection accepted to Bayesian Analysis! Joint work with Tony Canale (11/30/2023)
Bayesian Quantile Regression with Subset Selection: A Posterior Summarization Perspective submitted! Joint work with Joe Feldman (11/03/2023)
The projected dynamic linear model for time series on the sphere submitted! Joint work with John Zito (08/29/2023)
Bayesian adaptive and interpretable functional regression for exposure profiles accepted to Annals of Applied Statistics! Joint work with Yunan Gao (07/24/2023)
New documentation is now available for several R packages: countSTAR, SeBR, and BayesSubsets (07/17/2023)
Monte Carlo inference for semiparametric Bayesian regression submitted! Joint work with Bohan Wu (06/08/2023)
The R package SeBR implements these models and algorithms and is accompanied by detailed documentation and examples
Ultra-efficient MCMC for Bayesian longitudinal functional data analysis submitted! Joint work with Thomas Sun (06/05/2023)
R code is available on GitHub
Warped Dynamic Linear Models for Time Series of Counts accepted to Bayesian Analysis! Joint work with Brian King (06/05/2023)
The R package countSTAR is now available on CRAN! Check out the fantastic vignette developed by Brian King. Note that countSTAR replaces the previous GitHub package, rSTAR. (04/10/2023)
Congratulations to Joe Feldman on his James R. Thompson Student Award for “current students performing outstanding pre-thesis research judged on the basis of promise, publications, awards and other scholarly achievements” in Rice STAT! (01/30/2023)
Congratulations to Yunan Gao on her ASA Section on Bayesian Statistical Science student paper award! (01/15/2023)
Congratulations to Joe Feldman on his ASA Section on Bayesian Statistical Science and Survey Research Methods, Government Statistics, and Social Statistics Sections (declined) student paper awards! (01/15/2023)
Nonparametric Copula Models for Multivariate, Mixed, and Missing Data submitted! Joint work with Joe Feldman (10/26/2022)
R code is available on GitHub
Dynamic and Robust Bayesian Graphical Models accepted to Statistics and Computing! Joint work with Chunshan Liu and Marina Vannucci (10/26/2022)
Bayesian adaptive and interpretable functional regression for exposure profiles submitted! Joint work with Yunan Gao (10/05/2022)
R code is available on GitHub
Semiparametric discrete data regression with Monte Carlo inference and prediction submitted! Joint work with Bohan Wu (09/07/2022)
This is an updated version of “Conjugate priors for count and rounded data regression”
Racial residential segregation shapes relationships between early childhood lead exposure and 4th grade standardized test scores published in PNAS! Joint work with CEHI and affiliates (08/15/2022)
Adaptive dependent data models via graph-informed shrinkage and sparsity grant awarded by NSF! (08/04/2022)
Subset selection for linear mixed models accepted to Biometrics! (06/09/2022)
Bayesian subset selection and variable importance for interpretable prediction and classification accepted to Journal of Machine Learning Research! (04/20/2022)
Bayesian Data Synthesis and the Utility-Risk Trade-Off for Mixed Epidemiological Data accepted to Annals of Applied Statistics! Joint work with Joe Feldman (01/18/2022)
Congratulations to Brian King on his ASA Section on Bayesian Statistical Science and Business and Economic Statistics Section (declined) student paper awards! (01/15/2022)
Congratulations to Joe Feldman on his ASA Health Policy Statistics Section student paper award! (01/14/2022)
Semiparametric count data regression for self-reported mental health accepted to Biometrics! Joint work with Rice undergraduate Bohan Wu! (12/22/2021)
Congratulations to Brian King on his International Biometric Society ENAR Distinguished Student Paper Award! (12/17/2021)
Conjugate priors for count and rounded data regression submitted! (10/29/2021)
Warped Dynamic Linear Models for Time Series of Counts submitted! Joint work with Brian King (10/28/2021)
Received the Blackwell-Rosenbluth Award! (10/18/2021)
Received an Economics, Finance, and Business (EFaB) Presentation Award: European Seminar on Bayesian Econometrics! (09/03/2021)
Revised submission: Semiparametric Functional Factor Models with Bayesian Rank Selection, now with Tony Canale! (10/12/2021)
Subset selection for linear mixed models submitted! (07/26/2021)
Bayesian decision analysis for collecting nearly-optimal subsets uploaded as a spotlight talk for the ICML workshop SubSetML: Subset Selection in Machine Learning: From Theory to Practice (06/23/2021)
Semiparametric count data regression for self-reported mental health submitted! Joint work with Rice undergraduate Bohan Wu! (05/27/2021)
Bayesian Variable Selection for Understanding Mixtures in Environmental Exposures accepted to Statistics in Medicine! Joint work with Mercedes Bravo, Henry Leong, Alex Bui, Rob Griffin, Kathy Ensor, and Marie Lynn Miranda (05/26/2021)
Bayesian subset selection and variable importance for interpretable prediction and classification submitted! (04/16/2021)
Fast, Optimal, and Targeted Predictions using Parametrized Decision Analysis accepted to JASA! (02/11/2021)
A Bayesian Framework for Generation of Fully Synthetic Mixed Datasets submitted! Joint work with Joe Feldman (02/02/2021)
Congratulations to Chunshan Liu on her ASA Business and Economic Statistics Section Best Student Paper Award! (01/21/2021)
Semiparametric Functional Factor Models with Bayesian Rank Selection submitted! (09/08/2020)
Army Research Office (ARO) Young Investigator Award: “Optimal Bayesian Approximations for Targeted Prediction” (06/01/2020 - 09/30/2022)
New endowed chair: Dobelman Family Assistant Professor of Statistics (05/22/2020)
Rice University COVID grant: “Dynamic Functional Data Analysis: New Statistical Tools to Flatten the Curve" (Role: PI; co-PI Thomas Sun; 05/18/2020 - 08/18/2020)
Dynamic Regression Models for Time-Ordered Functional Data accepted to Bayesian Analysis! (05/05/2020)
R package available on GitHub: drkowal/dfosr
Elected Secretary/Treasurer 2021-2022 for the ASA Business and Economic Statistics Section! (05/02/2020)
Rice University COVID grant: “Air Quality Impacts of COVID Response Policies” (Role: co-PI; PI Daniel Cohan; 05/01/2020 - 12/31/2020)
Simultaneous transformation and rounding (STAR) models for integer-valued data published in EJS! Joint work with Tony Canale (03/25/2020)
Nonlinear state-space modeling approaches to real-time autonomous geosteering accepted to the Journal of Petroleum Science and Engineering! Joint work with Yinsen Miao, Nielkunal Panchal, Jeremy Villa, and Marina Vannucci (02/02/2020)
Bayesian Function-on-Scalars Regression for High Dimensional Data accepted to JCGS! Joint work with Daniel Bourgeois (12/12/2019)
Rice Data Science Conference: presented “Prediction Models for Integer and Count Data” (video link). Joint work with Tony Canale (10/14/2019)
Rice University Machine Learning Seminar: Presented “Prediction Models for Integer and Count Data”. Joint work with Tony Canale (9/25/2019)
Joint Statistical Meetings: Presented Bayesian Function-On-Scalars Regression for High-Dimensional Data in an invited session, “Modern and Practical Solutions to Difficult High-Dimensional Regression Problems”. Joint work with Daniel Bourgeois (7/31/2019)
Joint Statistical Meetings: Organized and chaired an invited session, “Modern Methods for Structured and Dynamically Dependent Data” (7/28/2019)
BNP12: Presented Dynamic Shrinkage Processes (6/21/2019)
Stochastic Clustering and Pattern Matching for Real Time Geosteering accepted to Geophysics! Joint work with Mingqi Wu, Yinsen Miao, Nielkunal Panchal, Marina Vannucci, Jeremy Villa, and Faming Liang (6/29/2019)
Integer-Valued Functional Data Analysis for Measles Forecasting accepted to Biometrics! (5/21/2019)
Dynamic Shrinkage Processes accepted to JRSS-B! (4/15/2019)
Daniel Bourgeois presented Bayesian Function-on-Scalars Regression for High Dimensional Data at COTS 2019 (4/5/2019 - 4/6/2019)
CMStatistics 2018: Presented Dynamic Function-on-Scalars Regression (12/14/2018 - 12/16/2018)
An R package for Bayesian Function-on-Scalars Regression for High Dimensional Data is available on GitHub! Joint work with Daniel Bourgeois (10/16/2018)
Install and load using the following code:
library(devtools)
devtools::install_github("drkowal/fosr")
library(fosr)
Uses include:
Bayesian estimation and inference for function-on-scalar regression: fosr(...)
Decision-theoretic approach to variable selection in functional regression: fosr_select()
Additional tools for plotting, simulations, and evaluation of model fit
European Seminar on Bayesian Econometrics (ESOBE): Presented Dynamic Function-on-Scalars Regression (10/11/2018 - 10/12/2018)
Daniel Bourgeois presented a poster, “Using Bayesian Posterior Summarization for Model Selection in Functional Data Analysis” at the Rice Data Science Conference (10/9/2019)
Joint Statistical Meetings: Presented Dynamic Shrinkage Processes in a special section for Business and Economic Statistics Section Best Student Paper Award winners and served as session chair for Analysis of Big Dynamically Dependent Data (7/29/2018 - 8/1/2018)
New Researchers Conference: Presented Dynamic Function-on-Scalars Regression and served as session chair (07/26/2018)
NBER-NSF Seminar on Bayesian Inference in Econometrics and Statistics: Presented Dynamic Shrinkage Processes (05/25/2018 - 05/26/2018)
Arnold Zellner Thesis Award in Econometrics and Statistics: Honorable Mention (Rice University story, 5/11/2018)
Rice University profile: "Dan Kowal: drawn to uncertainty"
Institute for Mathematics and its Applications: Forecasting from Complexity: Presented Dynamic Shrinkage Processes (04/23-2018 - 04/27-2018)
American Statistical Association, Houston Area Chapter: Presented Dynamic Shrinkage Processes (4/10/2018)
Presented Dynamic Shrinkage Processes at University of Washington, Department of Biostatistics (2/22/2018)
Presented Dynamic Shrinkage Processes at Baylor University, Department of Statistical Science (2/1/2018)
ASA Business and Economic Statistics Section Best Student Paper Award: Dynamic Shrinkage Processes (1/15/2018)
CMStatistics 2017 Presented Functional Autoregression for Sparsely Sampled Data in a special section on recent developments in functional time series analysis (12/16/2017 - 12/18/2017)
Joint Statistical Meetings: Presented Functional Autoregression for Sparsely Sampled Data in a special section for ASA Nonparametrics Section Student Paper Award winners and served as session chair for Bayes Theory and Foundations (7/31/2017 - 8/2/2017)
An R package for Dynamic Shrinkage Processes is available on GitHub! (7/25/2017)
Install and load using the following code:
library(devtools)
devtools::install_github("drkowal/dsp")
library(dsp)
Uses include:
Curve-fitting of irregular data via Bayesian trend filtering: btf(...)
An adaptive time-varying parameter regression model: btf_reg(...)
Curve-fitting of irregular data with unequally-spaced observations: btf_bpsline(...)
PhD dissertation submitted and approved: Bayesian Methods for Functional and Time Series Data, Cornell University, Department of Statistical Science (7/13/2017)
An R package for A Bayesian multivariate functional dynamic linear model is available on GitHub! (5/17/2017)
ASA Nonparametrics Section Student Paper Award: Functional autoregression for sparsely sampled data (3/16/2017)