Facebook MCS

Colloquium

The Department of Math and Computer Science presents a seminar series each semester. The department colloquium is traditionally held on Wednesdays at 4 pm and the Graduate Seminar is held on Mondays at 4 pm.  All Colloquia are open to the SDSM&T campus and also to the general public.

Check this page for dates, speakers, and topics.

Spring 2014

 

Stay Right Except to Pass

March 25 - McLaury 205, 11:00 AM

by Math Modeling Team: Christopher Cody, Erica Daniels, and Daniel Nix

 

The Mathematical Contest in Modeling was held last February and posed the problem of analyzing the advantage (if any) of the traffic rule that involves staying right except to pass. Students were to model this rule in various traffic patterns and compare this to other traffic rules that can be used to decide, which rules would be optimal. The Modeling Team will discuss the problem and present their results.  

 

 --------------------------------------------------------

 Many-core, Many More and Much Else

  March 18 - McLaury 205, 11:00 AM

 by Dr. Christer Karlsson

  

The combination of architectural and organizational enhancements led to 17 years of sustained annual growth rate of 50% from 1986-2003. The growth dropped to 22% per from 2003-210, mostly due to the combined promblem of power, memory and instruction-level parallelism. The hardware changes have significant impacts on software and applications. Commodity parallel hardware, such as multi-core CPUs, many core GPus and Intel MIC accelerators, provide enormous computing power which can potentially achieve significant speedups on existing geospatial processing and open the opportunities for new applications. this presentation is an introduction to what hardware and methods are available, and what advantages they might and might not have. We will show results from experiments using data-parallel design for point to  polyline shortest distance computation on commodity hardware, multi-core CPUs, GPUs and Vector Processing Units (VPU) that come with multi-core CPUs augmented with Intel MICs, all available at SDSM&T. We will compare their performance and discuss improtant factors that may significantly affect the performance.

---------------------------------------------------------------------

 

 Molecular Modeling and Simulation: Effective Tools for Chemical Engineering Design

March 4 - McLaury 205, 11:00 AM

by Dr. Kenneth Benjamin, Chemical and Biological Engineering

 

There exists a direct connection between the macroscopic behavior of an engineered system and the fundamental chemical and/or physical forces at the molecular level.  Ideally, chemical engineers would be able to exploit this molecular level information in their design, control, and optimization of macroscopic process performance.  Molecular modeling and simulation provides a mathematical and computational framework in which to make first-principles predictions of material properties and basic chemical/physical mechanisms.  This talk presents the basics of molecular modeling and simulation, as well as applications of these computational techniques towards chemical engineering problems in bioenergy, catalysis, and fuels.

---------------------------------------------------------- 

 


Establishing Property-Structure Relationships of Smart Materials through Molecular Modeling

March 4 - McLaury 205, 11:00 AM


by Dr. Kevin Hadley, Chemical and Biological Engineering

 

 Molecular modeling allows us to observe the dynamic behavior of materials at the molecular level.  These models are dictated by simple classical Newtonian physics; where, for example, bonds are represented by harmonic springs.  Through alteration of the physical parameters (force constant and equilibrium distance), the physics of the molecule can easily be altered within the model.  The goal of this research is to make systematic alterations to the base molecular model and observe the effects on macroscopic properties of interest like heat capacity and density or smart material behavior like self-healing or anti-icing. By “mapping” the parameter space, the target macroscopic behaviors can be observed by choosing the appropriate molecular level features (I.e. Bond stiffness).  This seminar is a summary of what has been explored and our future plans.

Assessment Tool Development

As education research is establishing itself as a more robust research area, the methods must become equally robust.  Specifically, when testing if an intervention strategy is effective, an observable change must be quantified to judge statistical significance.  This half of the talk will summarize a number of education research projects in need of robust assessment tools and general strategies for development.  

-------------------------------------------------------------------------------------

 

 

 

 Assessment as Learning: An experiment with abilities-based grading

February 4 - McLaury 205, 11:00 AM

by Dr. Travis Kowalski

Instructors face many, common difficulties in teaching introductory level mathematics courses, not least of which is how to encourage students to both work on mathematics and learn from their work. There are many different approaches to this, ranging from assigning daily homework to online assignments to regular quizzing to student podcasting; this presentation discusses a different approach, modelled after the “Assessment as Learning” principle employed by Alverno College.

 In this presentation, Dr. Kowalski will discuss his experiences experimenting with an “abilities-based, assessment as learning” grading system for his Fall 2013 Calculus courses. In it, he will outline his twin goals of the experiment and how he implemented it through gateway assignments and pass/fail content exams. It will conclude with a summary and breakdown of the end-of-semester results, together with reflections on the semester and improvements made for Spring 2014. 

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 Fall 2013


Undergraduate Talks

Tuesday, December 10 - McLaury 205, 11:00 AM

Christopher Cody

Jet-determinacy of solutions to a mixed composition equation: A Taylor Series approach

 An important problem in complex analysis is to find the power-series structure of solutions to equations involving the compositions of known and unknown analytic functions.  In this presentation, we investigate the structure of solutions H to the mixed composition equation H(Q(x)) = Q(H(x)) using Faa di Bruno’s Formula to isolate Taylor coefficients.  We develop sufficient conditions on Q to ensure that solutions exist and are uniquely determined by a finite jet, that is, by a set of initial conditions involving derivatives of a fixed order. 

Joshua Wipf

Third-order tensors acting on a unit sphere: A tool for visualization and analysis with applications in tensor comprehension

This presentation focuses on two related problems.  First, we implement a tool for visualization of tensors, specifically those related to the product of the unit sphere and the SVD (single value decomposition).  Second, we analyze how zero divisors interact with tensor multiplication. 

---------------------------------------------------------------------------------------------------------------------------- 

 Undergraduate talks

A day packed with reports from students finishing the first half of their math research project.

Wednesday, December 4 - McLaury 205, 4:00 PM

Erica Daniels - A Survival Analysis of the Hunger Games

Jessica Gillaspie - Introduction to Rocket Design

Eric Larsen - Numerical Approximations for the Polar Wave Equation

Anthony Morast - Isotopy Classes of Order 5 Latin Squares

Dean Laganiere - Verification of the Chemical Composition of a Drug

--------------------------------------------------------------------------------------------------- 

 Modeling Raindrop Size part 2

Dr. Roger Johnson
Wednesday, October 23 - McLaury 205, 4:00 PM

Characterizing raindrop size is useful in meteorology, hydrology, and telecommunications, among other fields. The particulars in uncovering an underlying raindrop size distribution, however, depend on the instrumentation available to record drop size. Part 1 examined the situation in which drop sizes were assumed to be accurately measured over the full range of drop sizes. In this presentation, Part 2, we look at the more realistic situation where this fails to be true.

--------------------------------------------------------   

Modeling Raindrop Size part 1
Dr. Roger Johnson
Wednesday,  October 16- McLaury 205, 4:00 PM

Characterizing raindrop size is useful in meteorology, hydrology, and telecommunications, among other fields. The particulars in uncovering an underlying raindrop size distribution, however, depend on the instrumentation available to record drop size. In this presentation, the first of two, we examine the situation in which drop sizes are accurately measured over the full range of drop sizes.

A small example involving the estimation of the number of German tanks during WWII (requiring mathematical background at the level of Math 381 or CSC 251) will be used to illustrate two basic statistical estimation procedures used in characterizing raindrop size.

This presentation is a partial summary of work with Dr. Donna Kliche and Dr. Paul Smith of the Institute for Atmospheric Sciences. Our fourth publication on characterizing raindrop size is in press.