Category Archives: Department of Statistics

LAS Gallery of Excellence

The first 25 selections for the 100 Years of LAS Gallery of Excellence (http://www.las.illinois.edu/100/excellence/).  This virtual gallery highlights some of the most prominent people and events throughout LAS history, which dates back to 1913.

Two notable statisticians/mathematicians made the list: Prof. Joseph Doob (the namesake of our department library) and Prof. David Blackwell.
++++++++++++++++++++++++++++++++++++++++

Dept of Statistics Seminar

Statistics Seminar: Sparse Principal Component Analysis: Optimal Rates and Adaptive Estimation
Speaker     Yihong Wu (University of Illinois)
Date           Oct 10, 2013
Time           4:00 pm – 5:15 pm
Location    106B8 Engineering Hall
Sponsor     Statistics Department
Event type Seminar
 
Abstract: Principal component analysis (PCA) is one of the most commonly used statistical procedures with a wide array of applications. This talk considers both minimax and adaptive estimation of the principal subspace in the high-dimensional setting. Under mild technical conditions, we first establish the optimal minimax rate of estimating the principal subspace which are sharp with respect to all parameters, thus providing a complete non-asymptotic characterization of the statistical difficulty of the problem. The rate-optimal estimator is constructed using aggregation, which, however, might not be computationally feasible. We then introduce an adaptive procedure for estimating the principal subspace which is fully data-driven and computationally efficient. It is shown that the estimator attains the minimax rate simultaneously over a large collection of the parameter spaces. A key idea in our construction is a scheme which reduces the sparse PCA problem to a high-dimensional regression problem. Time permitting, I will discuss optimal rank detection and the challenges in computational complexity.
+++++++++++++++++++++++++++++++++++++++++

Dept of Statistics Seminar

Statistics Seminar: Empirical and Sequential Empirical Copula Processes under Serial Dependence and Weak Smoothness Conditions
Speaker           Stanislav Volgushev (Ruhr-Universitaet Bochum)
Date                Oct 3, 2013
Time                4:00 pm – 5:15 pm
Location          106B8 Engineering Hall
Sponsor           Department of Statistics
 
The empirical copula process plays a central role for statistical inference on copulas. The main purpose of this talk is to give an overview of recent results on the asymptotic properties of this process. In the first part of the talk, we focus on process convergence with respect to the supremum norm. We also briefly discuss properties of sequential copula processes. It is known that in some interesting examples, weak convergence of the copula process with respect to the supremum norm does not hold. In the second part of the talk, we introduce a new metric that allows to handle most of those cases. Joint work with Axel Buecher and Johan Segers.
+++++++++++++++++++++++++++++++++

Grad School Preparation Month

October is Grad School Prep Month at The Career Center. We offer individual appointments to help students prepare for graduate school, review personal statements, and prepare for the application process. Additionally we will be providing a series of graduate school preparation events this month. Please advertise these events widely with students who are thinking about graduate school or applying this application season. Thank you!
 
Ø  Exploring Graduate School Workshop
o   Wednesday, October 9 at 4pm, The Career Center Conference Room 715 S Wright Street
o   Thursday, October 10 6pm, The Career Center Conference Room 715 S Wright Street
Ø  Graduate and Professional School Fair
o   Wednesday, October 16 1-5pm, Illini Union A,B,C
Ø  Graduate School Admissions Q &A Panel
o   Wednesday, October 16 5pm, Illini Union Colonial Room
Ø  Graduate School Application Process Workshop
o   Tuesday, October 22 6pm, The Career Center Conference Room 715 S Wright Street
Ø  Life in Graduate School- Panel with current graduate students
o   Wednesday, October 30 6pm, The Career Center Conference Room 715 S Wright Street
 
+++++++++++++++++++++++++++++++++++++++

Dept of Statistics Seminar

Asymptotically minimax empirical Bayes estimation of a sparse, Ryan Martin (University of Illinois, Chicago)
Speaker           Ryan Martin, (University of Illinois, Chicago)
Date                Sep 26, 2013
Time                4:00 pm – 5:15 pm
Location          106B8 Engineering Hall
Sponsor           Department of Statistics
Event type        Seminar
 
ABSTRACT: Estimating a sparse high-dimensional normal mean vector is an important classical problem. In this talk, I will introduce a new empirical Bayes model based on a unique data-dependent prior. I will show that, under some conditions, our empirical Bayes posterior distribution concentrates on balls, centered at the true mean vector, with squared radius proportional to the frequentist minimax rate for the given sparsity class. This result provides some new insight concerning the fully Bayes approach to this same problem. Asymptotic minimaxity of the corresponding empirical Bayes posterior mean is shown, and a simple Gibbs sampling algorithm for computation will be discussed. Finally, two simulation studies will be presented, demonstrating the strong finite-sample performance of the proposed estimator against a variety of popular alternatives. (This is joint work with Stephen Walker at the University of Texas at Austin.)
+++++++++++++++++++++++++++++++++++

TONIGHT: Illini Statistics First General Meeting

Please join Illini Statistics Club on Tuesday, September 17 at 5pm for our first event. It will be a picnic at Scott Park on 3rd and Healey. We will be serving food, passing out important information, and registering new members. Everyone is welcome!


Hope to see you there,

Illini Statistics Club
Department of Statistics
University of Illinois – Urbana Champaign
+++++++++++++++++++++++++++++++++++++++++++++++++++++

GRAD/UNDERGRAD PRESENTATION OPPORTUNITY

The American Association for the Advancement of Science (AAAS) is providing an opportunity for students to present their work in poster sessions at its annual meeting in Chicago February 13-17, 2014 (http://meetings.aaas.org/). Full-time undergraduate and graduate students are eligible to apply; if selected to present a poster, students will be able to have the meeting registration fee waived by serving as AAAS Meeting Session Aides and will be eligible to apply for travel funding.  Details are in the attached flyer; the deadline for submission of poster abstracts is October 24, 2013.

Students and their advisors are urged to take advantage of this unique opportunity to highlight statistical research and its role in interdisciplinary science at the annual meeting of the world’s largest general scientific society.  AAAS publishes the highly respected journal Science and is organized into 24 Sections representing all areas of science, including Section U (Statistics), which celebrated its 50th anniversary last year.

Thank you for your help.

Marie Davidian, President, American Statistical Association
Sally Morton, Chair, AAAS Section U (Statistics)

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Marie Davidian
2013 President, American Statistical Association
Professor
Department of Statistics
North Carolina State University

davidian@ncsu.edu
http://www4.stat.ncsu.edu/~davidian/

International Year of Statistics
www.statistics2013.org

Please send correspondence concerning Biometrics to biometrics@tibs.org.
(http://www.biometrics.tibs.org/)

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

What's Your Major? (Planet Money podcast)

Sure, some college degrees lead to higher paying jobs than others. But what’s shocking — at least, it was shocking to us — is just how big the gap can be.

The most lucrative majors typically lead to jobs with salaries over $100,000 a year. The least lucrative lead to salaries of around $30,000.

On today’s show, we run the numbers. We talk to people who majored in the most- and least-lucrative subjects. And we hear from an economist who says, when it comes to income, choosing a major is more important than choosing a college.

http://www.npr.org/blogs/money/2013/09/11/221417806/episode-485-whats-your-major

 

——————————————————————-

 

Anheuser-Busch lab taps into campus talent

The world’s largest brewer will be opening its first permanent residence on a university campus at the University Research Park.

Anheuser-Busch InBev will open its “Bud Lab,” a data analytics center, sometime this month.

Director Rafael Pinterich will oversee the lab, selecting and implementing real-world experimental problems and also working closely with students. Bud Lab researchers will focus on developing data research and innovation to find creative solutions to programs ranging from assortment optimization, social media and market trends to large scale data initiatives, according to a press release.
——————————————