David B. Dunson, Arts and Sciences Professor of Statistical Science and Mathematics and Faculty Network Member of Duke Institute for Brain Sciences

David B. Dunson

Development of novel approaches for representing and analyzing complex data.  A particular focus is on methods that incorporate geometric structure (both known and unknown) and on probabilistic approaches to characterize uncertainty.  In addition, a big interest is in scalable algorithms and in developing approaches with provable guarantees.

This fundamental work is directly motivated by applications in biomedical research, network data analysis, neuroscience, genomics, ecology, and criminal justice.   


Office Location:  218 Old Chemistry Bldg, Durham, NC 27708
Office Phone:  (919) 684-8025
Email Address: send me a message
Web Page:  https://github.com/david-dunson

Teaching (Fall 2018):

Teaching (Spring 2019):

Office Hours:

Thurs 9-10am
Education:

PhDEmory University1997
Ph.D.Emory University1997
B.S.Pennsylvania State University1994
Specialties:

Bayesian Statistics
Complex Hierarchical and Latent Variable Modelling
Nonparametric Statistical Modelling
Model Selection
Statistical Modeling
Research Interests: Nonparametric Bayes, Latent variable methods, Model uncertainty, Applications in epidemiology & genetics, Machine learning

Current projects: Nonparametric Bayes methods for conditional distributions, Semiparametric methods for high-dimensional predictors, New priors for functional data analysis, Borrowing information across disparate data sources, Methods for identifying gene x environmental interactions

Development of Bayesian methods motivated by applications with complex and high-dimensional data. A particular focus is on nonparametric Bayes approaches for conditional distributions and for flexible borrowing of information. I am also interested in methods for accommodating model uncertainty in hierarchical models, and in latent variable methods, including structural equation models. A recent interest has been in functional data analysis.

Areas of Interest:

Functional data analysis
Genetics
Latent variable methods
Machine learning
Molecular epidemiology
Nonparametric Bayes
Order restricted inference
Model selection and averaging

Keywords:

Action Potentials • Algorithms • Artificial Intelligence • B-Lymphocytes • Bayes Theorem • Computer Simulation • Data Interpretation, Statistical • DNA-Binding Proteins • Electrophysiological Phenomena • Epidemiologic Methods • Gene Dosage • Genetic Association Studies • Germinal Center • Longitudinal Studies • Markov Chains • Models, Statistical • Models, Theoretical • Monte Carlo Method • Multivariate Analysis • Neurons • Pattern Recognition, Automated • Phenotype • Probability • Stochastic differential equations

Representative Publications   (search)

  1. Dunson, DB, Nonparametric Bayes local partition models for random effects., Biometrika, vol. 96 no. 2 (2009), pp. 249-262, ISSN 0006-3444 [doi]  [abs]
  2. Bigelow, JL; Dunson, DB, Bayesian semiparametric joint models for functional predictors, Journal of the American Statistical Association, vol. 104 no. 485 (2009), pp. 26-36, ISSN 0162-1459 [doi]  [abs]
  3. Dunson, DB; Xing, C, Nonparametric Bayes Modeling of Multivariate Categorical Data., Journal of the American Statistical Association, vol. 104 no. 487 (2009), pp. 1042-1051, ISSN 0162-1459 [doi]  [abs]
  4. Dunson, DB; Park, J-H, Kernel stick-breaking processes, Biometrika, vol. 95 no. 2 (2008), pp. 307-323, ISSN 0006-3444 [doi]  [abs]
  5. Dunson, DB; Herring, AH; Engel, SM, Bayesian selection and clustering of polymorphisms in functionally related genes, Journal of the American Statistical Association, vol. 103 no. 482 (2008), pp. 534-546, ISSN 0162-1459 [doi]  [abs]
Recent Grant Support