Opening meeting of the CRG


This meeting will kick off a series of activities which will be held in the pacific northwest (PNW) as part of a recently awarded PIMS collaborative research group (CRG) on Bayesian modeling and computation for networks.The goal of this opening meeting is bring together researchers working on Bayesian modeling for networks in the PNW from different communities, thereby fostering collaborations and intellectual exchange. The meeting will open with Keynote presentation given by a prominent researcher in the field. The remainder of the meeting will be spent with presentations on a variety of research projects on networks currently undertaken in the PNW. Finally, the workshop will end by a discussion of the future plans for our CRG. The organization of the meeting has been specifically designed to encourage the graduate students attending the meeting to interact among themselves and with the other researchers. Each student attending the meeting will be expected to present his/her work in a short talk or in a poster. There will be ample time for discussions that will give the students great opportunities to receive feedback about their work, ask additional questions and to establish new connections.

Location: Semiahmoo Resort

Date: May 3-4, 2008

Organizers: Adrian Dobra, Raphael Gottardo, Lurdes Inoue

Sponsors: PIMS, VIGRE

Keynote Speaker: Prof. Michael Jordan

Important Dates:
Deadline for participation confirmation, March 15, 2008.
Deadline for title/abstract submission, March 31, 2008. Titles/Abstracts should be sent to Raphael Gottardo. When you submit your abstract, please let me know if you'd prefer to give an oral or poster presentation.


Confirmed participants (in no particular order):
Faculty members/Postocs: Michael Jordan, Adrian Raftery, Peter Hoff, Tim Swartz, Kevin Murphy, Raphael Gottardo, Adrian Dobra, Paramjit Gill, Lurdes Inoue, Nando de Freitas, Ka Yee Yeung, Dave Campbell, Dale Schuurmans, Paul Gustafson, Donatello Telesca (PDF), Francois Caron (PDF).
Students: Le Bao, Mark Schmidt, Emtiyaz Khan, Aline Tabet, Leontine Alkema, Alex Lenkoski, Saman Muthukumarana, Xiaoyue Niu, Xiaohui Chen, Pavel Krivitsky, Ranran Wang, Ryan Admiraal, Youyi Fong, Guillaume Alain, and Hoyt Koepke. Note that students will be expected to share a room (2 students/room).

Preliminary program (may change slightly):

Friday
May 2
Saturday
May 3
Sunday
May 4
8:00-9:00 Breakfast Breakfast
9:00-10:00 Keynote Lecture
Michael Jordan
Talks: Session 3
10:00-10:30 Coffee break Coffee break
10:30-12:00 Talks: Session 1 Talks: Session 4
12:00-1:30 Lunch Lunch
1:30-3:00 Talks: Session 2 Departure
3:30-6:00 Arrival/ Check-in Coffee break
Free time
6:00-8:00 Arrival/ Check-in Dinner
8:00-11:00 Poster Session

Keynote Lecture

Speaker: Michael I. Jordan
Department of Statistics, Department of Electrical Engineering and Computer Science
University of California, Berkeley

Title: "Nonparametric Bayesian Graphical Models"

Most of the development of graphical models has taken place within the realm of parametric statistics---the distributions that have been considered are the multinomial, the Gaussian and other distributions in the exponential family, and the set of graphs under consideration in a given problem is generally taken to be fixed and finite.  In this talk I overview some of the stochastic processes that are allowing us to move beyond these classical parametric restrictions; these include the Dirichlet process, the beta process, the gamma process, and various recursive constructions (e.g., Bayesian hierarchies) involving these processes.  I discuss some of the marginals that arise from integrating out these stochastic processes; these are guaranteed to enhance your appetite for nonparametrics.  I also present applications of models based on these stochastic processes
to problems in structural biology, statistical genetics, natural language processing and computational vision.

[Joint work with Yee Whye Teh and Romain Thibaux.]

Talks and poster titles:


Talks should be 30 minutes long including questions.

Session 1:
1) Speaker: Pavel Kritvitski
    Title: "Extensions and applications of latent cluster models for social networks"
2) Speaker: Xiaoyue Niu
     Title: "A Hidden Markov Model with Latent Eigen-factors for Dynamic Social Networks"
3) Speaker: Donatello Telesca
     Title: "Modeling Dependent Gene Expression"

Session 2:
1) Speaker: Francois Caron
    Title: "Clustering of sparse linear regression"
2) Speaker: Mark Schmidt
    Title: "Efficient Optimization Algorithms for Group-Sparsity"
3) Speaker: Alex Lenkoski
    Title: "Nondecomposable Gaussian Graphical Models and the Mode Oriented Stochastic Search"

Session 3:
1) Speaker: Leontine Alkema
    Title: "Uncertainty in fertility estimates and predictions"
2) Speaker: Saman Muthukumarana
    Title: "A Bayesian Model for One-day International Cricket Matches"

Session 4:
1) Speaker: Thordis Linda Thorarinsdottir
     Title: "Probabilistic wind forecasting using truncated normal regression over an
ensemble of forecasts"
2) Speaker: Le Bao
    Title: "Incremental Mixture Importance Sampling and its Application"
3) Speaker: Dave Campbell
    Title: "MCMC Methods for Dynamic Systems Models"


Poster session:
1) Speaker: Mohamma Emtiyaz Khan
   Title: "A comparison of algorithms for variable selection"
2) Speaker: Guillaume Alain
   Title: "Improved dependency net structure learning"
3) Speaker: Xioahui Chen
   Title: "Sparse Bayes regression with a solution via MM algorithm"
4) Speaker: Youyi Fong
   Title: "Comparing Two Hidden 2-Compartment Models for Hematopoiesis"
5) Speaker: Ryan Admiraal
   Title: "Sequential importance sampling for bipartite graphs"
6) Speaker: Ranran Wang
   Title: "Modeling Homophily of Research/Social Interactions of Graduate Students"
7) Speaker: Hoyt Koepke
   Title: "Bayesian Clustering Validation"