Spring 2014 and Fall 2013
Time and place: 3:45pm--4:45pm, Fridays at MSC 459; Cookies available in the lounge at 3:30pm.
April 25, 2014
Dr. Marie Snipes, Kenyon College
April 18, 2014
Dr. Alan Nathan, University of Illinois
April 11, 2014
Dr. Alessandro Ottazzi, Fondazione Bruno Kessler
March 28, 2014
Dr. Rui Feng, University of Pennsylvania
March 21, 2014
Dr. Sharon Senk, Michigan State University
February 28, 2014
Dr. Dilum P De Silva, Bowling Green State University at Firelands
November 15, 2013
Dr. Sean Li, University of Chicago
Title: Coarse differentiation of Lipschitz functions
Abstract: Bates, Johnson, Lindenstrauss, Preiss, and Schechtman introduced a notion of large scale differentiation. We discuss an extension of this to the nonabelian setting of Lipschitz maps between Carnot groups and it's application in quantitative nonembeddability.
November 8, 2013
Dr. Dan Burghelea, Ohio State University
Title: Data Analysis, Persistent Homology and Computational Morse Theory
Abstract: I will explain how topology/geometry, via ideas from Morse theory, proposes to bring some light in often very unstructured and large amount of data we get addicted to collect. I will also indicate how this process suggests new mathematics
November 1, 2013
Dr. Deane Arganbright, Divine Word University (Papua New Guinea)
Title: Creative Mathematical Visualization, Modeling, and Interdisciplinary Applications via Spreadsheets
Abstract: A spreadsheet, such as Microsoft Excel, is a creative instrument for the mathematical disciplines. In the classroom, it closely fits the way in which we typically do mathematics, while giving students experience with the principal mathematical tool of the workplace. We can implement many mathematical concepts and algorithms in Excel in a natural manner. This talk presents a diverse range of animated modeling and visualization applications developed during 30 years of experience in universities around the world. We also provide creative interdisciplinary and multicultural uses, such as using mathematics to create pictorial alphabet books for the national languages of Papua New Guinea. While the talk will provide new ideas for mathematicians, others will also find it readily accessible.
(Additional information: Dr. Arganbright is a 1962 B.S. graduate of BGSU, with a 1967 PhD in finite groups from the U. of Washington. He has held positions at Iowa State U., Whitworth U., and U. of Tennessee at Martin, as well as overseas positions at the U. of Papua New Guinea, Bendigo CAE (Australia), U. of Vienna, KAIST (South Korea), and Divine Word U. (PNG). He has been a pioneer in developing mathematical uses of spreadsheets, and has given presentations in over 25 counties, including 2 MAA Invited hour addresses. Additionally, he has published books and journal articles in this area.)
October 25, 2013
Dr. John P. Nolan, Math/Stat Department, American University, Washongton, D.C.
Title: Stable distributions: models for heavy tailed data
Abstract: Stable distributions are a class of heavy tailed probability distributions that generalize the Gaussian distribution and that can be used to model a variety of problems. An overview of univariate stable laws is given, with emphasis on the practical aspects of working with stable distributions. Then a range of statistical applications will be explored. If there is time, a brief introduction to multivariate stable distributions will be given.
October 18, 2013
Dr. Daniel Farley, Department of Mathematics, Miami University
Title: The lower algebraic K-theory of Hilbert modular groups
October 4, 2013
Dr. Kaibo Wang, Tsing Hua University
Title: Engineering Knowledge Driven Statistical Modeling for Spatial Data
Abstract: In certain complex manufacturing systems, the quality of a product is adequately characterized by a high-dimensional data map rather than by single or multiple variables. Such data maps also preserve unique spatial structures. Therefore, variation pattern analysis and statistical modeling based on the data map become very important for enhanced process understanding and quality improvement.
Using a real wafer example from semiconductor manufacturing and a carbon nano tube example from nano-manufacturing, we demonstrate how statistical models can be developed by incorporating engineering knowledge. In the wafer example, a three-stage hierarchical model is proposed. The wafer surface variation is decomposed into the macro- and micro-scale variations, which are modeled as a cubic curve and a first-order intrinsic Gaussian Markov random field, respectively. In the carbon nano tube example, a piece-wise polynomial model with spatial auto-regressive disturbance is developed. These examples show that engineering knowledge driven statistical modeling can play an important role in quality control of complex systems, and is also a promising area for statistical research.
September 27, 2013
Dr. Martin Mohlenkamp, Department of Mathematics, Ohio University
Title: If the Multiparticle Schrodinger Equation were easy to solve, then Chemistry would be too boring to support life.
Abstract: The multiparticle Schrodinger equation is the basic governing equation in quantum mechanics. Many person-centuries and cpu-millennia have been spent constructing approximate solutions to it. We should be glad it is so hard to solve because its subtle behavior allows the rich Chemistry upon which life depends.
I will describe the multiparticle Schrodinger equation and explain (some of) the reasons it is difficult to solve:
high-dimensionality, antisymmetry, scaling to large systems, inter-particle cusps, singular potentials and nuclear cusps, odd function spaces, etc. I will also describe our efforts to overcome these difficulties.
September 13, 2013
Dr. Artem Zvavitch, Department of Mathematical Sciences, Kent State University
September 6, 2013
Dr. Robert Gramacy, School of Business, University of Chicago
Title: Estimating Player Contribution in Hockey with Regularized Logistic Regression
Abstract: We present a regularized logistic regression model for evaluating player contributions in hockey. The traditional metric for this purpose is the plus-minus statistic, which allocates a single unit of credit (for or against) to each player on the ice for a goal. However, plus-minus scores measure only the marginal effect of players, do not account for sample size, and provide a very noisy estimate of performance. We investigate a related regression problem: what does each player on the ice contribute, beyond aggregate team performance and other factors, to the odds that a given goal was scored by their team? Due to the large-p (number of players) and imbalanced design setting of hockey analysis, a major part of our contribution is a careful treatment of prior shrinkage in model estimation. We showcase two recently developed techniques -- for posterior maximization or simulation -- that make such analysis feasible. Each approach is accompanied with publicly available software and we include the simple commands used in our analysis. Our results show that most players do not stand out as measurably strong (positive or negative) contributors. This allows the stars to really shine, reveals diamonds in the rough overlooked by earlier analyses, and argues that some of the highest paid players in the league are not making contributions worth their expense.