2013-14: LDHD: Statistical Inference in Sparse High-dimensional Models: theoretical and computational challenges: February 24-26, 2014

Workshop Information

This workshop was held at the Hamner Conference Center at the NC Biotechnology Center, 15 TW Alexander Drive, Research Triangle Park, NC.

This workshop focused on both theoretical and computational developments in high-dimensional statistical models. Of particular interest were models that involve high-dimensional matrix estimation, such as elliptical copula models, graphical and network models, factor models, and functional data. These models are typically parametrized by matrices of reduced complexity, for instance of low rank, low effective rank, with sparse patterns, or some combination of these. The low-complexity assumptions are crucial for the successful implementation and theoretical analysis of such models, especially from a limited amount of data.

High-dimensional models with low-dimensional structures are ubiquitous. Rich applications occur in genetics, neuroscience, economics, public health, psychology and sociology. New scientific challenges in these established areas, or in emerging areas such as medical geology or action science, arise on a continual basis, and with them the need to meet them at both computational and theoretical levels.

This workshop brought together researchers in applied, computational, and theoretical statistics, with the goals of (i) identifying pressing scientific open questions that can be answered within the framework of the workshop; (ii) disseminating state of the art results in the area of high dimensional statistical inference; and (iii) identifying open theoretical and computational challenges in this area.

Schedule
Participant List
Speaker Titles and Abstracts
Poster Titles

Schedule

Monday, February 24, 2014
The Hamner Conference Center Auditorium at the NC Biotechnology Center

8:30-8:55 a.m. Registration
8:55-9:00 Welcome
9:00-9:45 Alexandre Tsybakov, CREST-ENSAE
Linear and Conic Programming Approaches to High-Dimensional Errors-in-variables Models
9:45-10:00 Break
10:00-10:45 Cun-Hui Zhang, Rutgers University
Graphlet Screening in High Dimensional Variable Selection
10:45-11:00 Break
11:00-11:45 Jacob Bien, Cornell University
Convex Banding of the Covariance Matrix
11:45-1:00 Lunch (provided in the Congressional Room at the Hamner Conference Center)
1:00-1:45 George Michailidis, University of Michigan
Change Point Inference in Dynamic Erdos-Renyi Random Graphs
1:45-2:00 Break
2:00-2:45 Han Liu, Princeton University
Transelliptical Modeling and its Applications
2:45-3:00 Break
3:00-3:45 Luo Xiao, Johns Hopkins University
Estimation of Covariance Matrices with Particular Structures
3:45-4:00 Break
4:00-4:45 Xiaotong Shen, University of Minnesota
Ordinal Classification with Unstructured Predictors
4:45-5:00 Break
5:00-6:30 Poster Session and Reception (Galleria)

SAMSI will provide poster presentation boards and tape. The board dimensions are 4 ft. wide by 3 ft. high. They are tri-fold with each side being 1 ft. wide and the center 2 ft. wide. Please make sure your poster fits the board. The boards can accommodate up to 16 pages of paper measuring 8.5 inches by 11 inches.

#mce_temp_url#

Tuesday, February 25, 2014
The Hamner Conference Center Auditorium at the NC Biotechnology Center

 

9:00-9:45 Tony Cai, University of Pennsylvania
Recovery of High-Dimensional Low-Rank Matrices
9:45-10:00 Break
10:00-10:45 Harrison Zhou, Yale University
Rate-Optimal Posterior Contraction for Sparse PCA
10:45-11:00 Break
11:00-11:45 Venkat Chandrasekaran, Cal Tech
Computational and Statistical Tradeoffs via Convex Relaxation
11:45-1:00 Lunch (provided in the Congressional Room at the Hamner Conference Center)
1:00-1:45 Sofia Olhede, University College London
Nonparametric Graphon Estimation
1:45-2:00 Break
2:00-2:45 Yannick Baraud, University of Nice
ρ-estimation
2:45-3:00 Break
3:00-3:45 Christophe Giraud, Université Paris-Sud
On Estimator Selection
3:45-4:00 Break
4:00-4:45 Hao Helen Zhang, University of Arizona
Selection of Interaction Effects for Ultra High-Dimensional Data

Wednesday, February 26, 2014
The Hamner Conference Center Auditorium at the NC Biotechnology Center

 

9:00-9:45 Chris Holmes, Oxford University
Computational Decision Theory and Bayesian Methods for Exploring Sparse Structural Aberrations in Cancer Genomes
9:45-10:00 Break
10:00-10:45 Andrew Nobel, University of North Carolina
Hypothesis Testing and Community Detection
10:45-11:00 Break
11:00-11:45 Cosma Shalizi, Carnegie Mellon University
11:45-1:00 Lunch (provided in the Congressional Room at the Hamner Conference Center)
1:00-3:00 Post-Workshop Discussion

clickhereornot