Organizing Committee
Abstract

Data-rich investigations need advanced tools for allowing data to inform and interact with models. Bayesian Nonparametrics is a rapidly growing subfield of statistics and machine learning that provides a framework for creating complex statistical models that are both expressive and tractable. Recent, successful applications of nonparametric Bayesian models across a variety of domains suggests that these models have the potential for wide use. The challenge of constructing and using models on very high dimensional or even infinite dimensional spaces creates many opportunities for fruitful interactions between mathematicians, statisticians and computer scientists. Areas of interest include prior construction, posterior inference, posterior asymptotics, algorithmic development, and practical applications.

Image for "Bayesian Nonparametrics"
Image courtesy of Daniel Roy and Yee Whye Teh

Confirmed Speakers & Participants

  • Speaker
  • Poster Presenter
  • Attendee
  • Virtual Attendee

Workshop Schedule

Monday, September 17, 2012
TimeEventLocationMaterials
8:30 - 8:55am EDTRegistration & Welcome Coffee11th Floor Collaborative Space 
8:55 - 9:00am EDTWelcome - ICERM Director, Jill Pipher11th Floor Lecture Hall 
9:15 - 10:00am EDTUsing Bayesian Nonparametrics for Practical Optimization of Machine Learning Algorithms - Ryan Adams, Harvard University11th Floor Lecture Hall 
10:00 - 10:30am EDTCoffee/Tea Break11th Floor Collaborative Space 
10:30 - 11:15am EDTComputer Emulation with Non-Stationary Gaussian Processes - Surya Tokdar, Duke University11th Floor Lecture Hall
11:15 - 12:00pm EDTBayesian nonparametrics for adaptive surveys and tests - Lawrence Carin, Duke University11th Floor Lecture Hall 
12:00 - 2:00pm EDTBreak for Lunch  
2:00 - 2:45pm EDTConvergence of latent mixing measures in finite and infinite mixture models - Long Nguyen, University of Michigan11th Floor Lecture Hall
2:45 - 3:30pm EDTDirichlet process mixtures are inconsistent for the number of components in a finite mixture - Jeffrey Miller, Brown University11th Floor Lecture Hall
3:30 - 4:00pm EDTCoffee/Tea Break11th Floor Collaborative Space 
4:00 - 4:45pm EDTBayesian estimation of the discrepancy with misspecified parametric models - Pierpaolo De Blasi, Universita di Torino11th Floor Lecture Hall
5:00 - 6:30pm EDTWelcome Reception11th Floor Collaborative Space 
Tuesday, September 18, 2012
TimeEventLocationMaterials
9:15 - 10:00am EDTBayesian Nonparametric Approaches for Partially-Observable Reinforcement Learning - Finale Doshi-Velez, Massachusetts Institute of Technology11th Floor Lecture Hall 
10:00 - 10:30am EDTCoffee/Tea Break11th Floor Collaborative Space 
10:30 - 11:15am EDTSpecies sampling priors for the analysis of array CGH data - Michele Guindani, University of Texas, MD Anderson Cancer Center11th Floor Lecture Hall 
11:15 - 12:00pm EDTOn a simple class of Bayesian nonparametric autoregression models - Fernando Quintana, Pontifical Catholic University of Chile11th Floor Lecture Hall
12:00 - 5:00pm EDTAfternoon Free for Collaborations  
Wednesday, September 19, 2012
TimeEventLocationMaterials
9:15 - 10:00am EDTNonparametric Bayes tensor factorizations for big data - David B. Dunson, Duke University11th Floor Lecture Hall
10:00 - 10:05am EDTGroup Photo11th Floor Lecture Hall 
10:05 - 10:30am EDTCoffee/Tea Break11th Floor Collaborative Space 
10:30 - 11:15am EDTBayesian Emulation and Calibration of a Dynamic Model for an H1N1 Influenza Epidemic - Marian Farah, MRC Biostatistics Unit11th Floor Lecture Hall
11:15 - 12:00pm EDTDiscovery and prediction from clinical physiologic data - Suchi Saria, Stanford University11th Floor Lecture Hall 
12:00 - 2:00pm EDTBreak for Lunch  
2:00 - 2:45pm EDTNonparametric priors for exchangeable graphs and arrays - Peter Orbanz, Columbia University11th Floor Lecture Hall
2:45 - 3:30pm EDTFurther structure underlying the beta (and Indian buffet) process - Daniel Roy, University of Cambridge11th Floor Lecture Hall
3:30 - 4:00pm EDTCoffee/Tea Break11th Floor Collaborative Space 
4:00 - 4:45pm EDTMultiresolution Gaussian Processes - Emily Fox, University of Washington11th Floor Lecture Hall
7:00 - 8:30pm EDTPoster Session & Dessert Reception11th Floor Collaborative Space and Lecture Hall 
Thursday, September 20, 2012
TimeEventLocationMaterials
9:15 - 10:00am EDTClusters and features from combinatorial stochastic processes - Tamara Broderick, University of California, Berkeley11th Floor Lecture Hall
10:00 - 10:30am EDTCoffee/Tea Break11th Floor Collaborative Space 
10:30 - 11:15am EDTSlice sampling dependent normalized random measures - Sinead Williamson, Carnegie Mellon University11th Floor Lecture Hall 
11:15 - 12:00pm EDTThe tilted Beta process - Nils Lid Hjort, University of Oslo11th Floor Lecture Hall
12:00 - 12:05pm EDTSurvey Distribution  
12:05 - 2:00pm EDTBreak for Lunch  
2:00 - 2:45pm EDTNonparametric priors for analyzing censored data and data from repair models - Jayaram Sethuraman, Florida State University11th Floor Lecture Hall
2:45 - 3:30pm EDTCanonically correlated random measures and an application to survival analysis - Antonio Lijoi, Universita di Pavia11th Floor Lecture Hall
3:30 - 4:00pm EDTCoffee/Tea Break11th Floor Collaborative Space 
4:00 - 4:45pm EDTA nonparametric regression model for Bayesian survival analysis - Chris Holmes, University of Oxford11th Floor Lecture Hall
Friday, September 21, 2012
TimeEventLocationMaterials
9:15 - 10:00am EDTBayes factors for high and infinite dimensional models - Steve MacEachern, Ohio State University11th Floor Lecture Hall
10:00 - 10:30am EDTCoffee/Tea Break11th Floor Collaborative Space 
10:30 - 11:15am EDTOn adaptation for the posterior distribution under local and sup-norm losses - Judith Rousseau, Universite de Paris-Dauphine11th Floor Lecture Hall
11:15 - 12:00pm EDTPosterior Consistency in Kullback-Leibler Divergence with Applications to Misspecified Infinite Dimensional Models - Andriy Norets, Princeton University11th Floor Lecture Hall 
12:00 - 2:00pm EDTBreak for Lunch  
2:00 - 2:45pm EDTModeling exchangeable collections of networks using nonparametric Bayesian mixtures - Abel Rodriguez, University of California, Santa Cruz11th Floor Lecture Hall 
2:45 - 3:30pm EDTRandom Partition Distributions Indexed by Pairwise Information - David Dahl, Brigham Young University11th Floor Lecture Hall
3:30 - 4:00pm EDTCoffee/Tea Break11th Floor Collaborative Space 
4:00 - 4:45pm EDTOn some distributional properties of Gibbs-type priors - Igor Pruenster, University of Torino & Collegio Carlo Alberto11th Floor Lecture Hall
5:00 - 6:00pm EDTWhat do rough paths have to do with finance - Terry Lyons, University of Oxford11th Floor Lecture Hall 

Tutorial Week Schedule

Thursday, September 13, 2012
TimeEventLocationMaterials
9:00 - 10:30am EDTGaussian Processes I - Ryan Adams, Harvard University11th Floor Lecture Hall 
10:30 - 10:50am EDTCoffee/Tea Break11th Floor Collaborative Space 
10:50 - 12:20pm EDTGaussian Processes II - Ryan Adams, Harvard University11th Floor Lecture Hall 
12:20 - 1:40pm EDTBreak for Lunch  
1:40 - 3:10pm EDTCombinatorial Stochastic Processes I - Tamara Broderick, University of California, Berkeley11th Floor Lecture Hall 
3:10 - 3:30pm EDTCoffee/Tea Break11th Floor Collaborative Space  
3:30 - 5:00pm EDTCombinatorial Stochastic Processes II - Tamara Broderick, University of California, Berkeley11th Floor Lecture Hall 
Friday, September 14, 2012
TimeEventLocationMaterials
9:30 - 10:30am EDTMatlab Practical: Gaussian Processes - Dilan Görür, University of California, Irvine11th Floor Lecture Hall 
10:30 - 11:00am EDTCoffee/Tea Break11th Floor Collaborative Space 
11:00 - 12:00pm EDTMatlab Practical: Dirichlet Processes - Dilan Görür, University of California, Irvine11th Floor Lecture Hall 
12:00 - 1:30pm EDTBreak for Lunch  
1:30 - 3:30pm EDTTheoretical Foundations I - Peter Orbanz, University of Cambridge11th Floor Lecture Hall 
3:30 - 4:00pm EDTCoffee/Tea Break11th Floor Collaborative Space 

Associated Semester Workshops

Computational Challenges in Probability
Image for "Computational Challenges in Probability"
Uncertainty Quantification
Image for "Uncertainty Quantification"
Performance Analysis of Monte Carlo Methods
Image for "Performance Analysis of Monte Carlo Methods"

Lecture Videos

On some distributional properties of Gibbs-type priors

Igor Pruenster
University of Torino & Collegio Carlo Alberto
September 21, 2012

Bayes factors for high and infinite dimensional models

Steve MacEachern
Ohio State University
September 21, 2012

The tilted Beta process

Nils Lid Hjort
University of Oslo
September 20, 2012

Slice sampling dependent normalized random measures

Sinead Williamson
Carnegie Mellon University
September 20, 2012

Clusters and features from combinatorial stochastic processes

Tamara Broderick
University of California, Berkeley
September 20, 2012

Multiresolution Gaussian Processes

Emily Fox
University of Washington
September 19, 2012

On a simple class of Bayesian nonparametric autoregression models

Fernando Quintana
Pontifical Catholic University of Chile
September 18, 2012

Bayesian Nonparametric Approaches for Partially-Observable Reinforcement Learning

Finale Doshi-Velez
Massachusetts Institute of Technology
September 18, 2012