Abstracts

 

Eastern Kentucky University’s 19th Annual Symposium in

Mathematical, Statistical and Computer Sciences

Kennamer Room, Powell building

 

9:05

 

Comparing the Success of

Developmental to Non-Developmental Students in College Credit-Bearing Courses

 

Karen Sue Cain, Director

Developmental Education and Academic Testing

Assistant Professor Mathematics and Statistics

Eastern Kentucky University

 

Abstract

A study was conducted to determine how developmental education students at Eastern Kentucky University were progressing in college level courses.  Success rates in credit bearing courses for English, mathematics and reading intensive courses were calculated for students entering the university with developmental needs and students entering without developmental needs.  The results were wonderful.   Developmental students progressed at nearly the same rate as students entering the university without developmental needs! 

 

This session will focus on three areas: methodology, a comparison of success rates, and why only small differences exist between the two groups.  More specifically, the session will provide a methodology for comparing developmental to non-developmental students in “college level courses” in each content area, a comparison of success rates for students with and without developmental  needs in credit-bearing English, mathematics and reading intensive courses, and an explanation of why our developmental program is able to demonstrate a way to prepare students for college level courses

 

Faculty

 

 

9:25

 

  Surveys: Snail-Mail versus E-Mail

 

Sanjita Thapa-Chhetri,

Eastern Kentucky University

Mentor: Dr. Patti Costello

 

 

Research Objectives:

i)          Determine approximately how many EKU students already have a laptop computer, and how many have a desktop computer but who could have purchased a laptop instead,

ii)         Compare the response rate of an online survey and a mail-in survey.

 

We are using two methods to conduct our survey, bulk mailing through the University mail system and an electronic survey through e-mail. Random samples of full-time and part-time undergraduate students were asked to fill out either an online survey or a survey sent to them through the regular mail.

 

Undergraduate student

9:45

 

Construction of Fractions in Geometer's Sketchpad

 

Stephen Greene

Eastern Kentucky University

 

Abstract

Using Geometer’s Sketchpad to develop a new geometric construction and using that construction to get an infinite series of odd denominator fractions of a segment.

 

Undergraduate student

 

 

10:05

 

Panoramic image stitching

Daniel Chong1, Maneesh Kumar1, Peterkin Quagraine1*, Valentina Staneva1 and Selim Esedoglu2.

 

Abstract

To capture a panoramic view, expensive tools are needed such as highly specialized cameras with fish-eye lenses. Panoramic image stitching, the problem of stitching together multiple overlapping images of the same scene, affords us the opportunity to capture panoramic views using a normal camera. Applications of this include aerial photography, medical imaging, and virtual scene reconstruction. This talk would describe an approach to panoramic image stitching with emphasis on automatic registration, including procedures in current literature and modifications to these methods.

1. Summer Research Student, University of California at Los Angeles, Institute for Pure and Applied Mathematics, Research in Industrial Projects for the Summer 2004

2. Assistant Professor, Mathematics Department, University of California at Los Angeles

Undergraduate student, Berea College

 

 

 

10:35

 

Algorithms for Finding Short Addition Chains

 

Nathan Gilbert* & Duane Skaggs

Morehead State University

 

Abstract

Addition chains provide an opportunity to speed up exponentiation by reducing the number of 'slow' multiplications needed. This presentation will discuss what an addition chain is and describe some

techniques for generating short chains.

 

Undergraduate student

 

 

 

 

10:55

 

Technology time warp: Interning in rural Appalachia

 

Sarah Clayton &Bethany Loftis

Berea College

 

Abstract

Many local governments in rural Appalachia are less technologically advanced than one would hope for in the 21st century.  While the pride of some Appalachians tends to prohibit remedying these problems, Letcher County is much more forward thinking and employed four interns to update their current system.  Discussed in this presentation will be problems faced and overcome by those interns in regards to technological issues and the mindset of rural Appalachia.

Undergraduate students

 

11:15

 

 

An Overview of Quantum Computing

Rus May

Morehead State University

 

 

Abstract

Quantum computing is the study of computations that quantum mechanical systems can perform. By exploiting effects of parallelism, quantum algorithms can be vastly more efficient than their classical counterparts. To gain an overview, we compare machines based on quantum mechanics with those based on von Neumann architecture and then investigate the differences in implementing algorithms in both paradigms. One example of an algorithm that is exponentially more efficient on a quantum machine is factoring integers. This algorithm is important since it forms the base for security of many Internet transactions. We also look at quantum machines that have been built in laboratories, their commercial viability, and some simulators. There are surprisingly few prerequisites for the subject, although a background in classical algorithms and linear algebra is useful. This talk, however, is a gentle introduction, accessible to students and faculty alike.

 

 

 

12:05

 

Lunch

 

 

 

1:30

On unavoidable objects

 

Jenö Lehel

The University of of Memphis and

Computer and Automation Research Institute of the Hungarian Academy of Sciences

 

 

 

2:30

 

Linux vs. Windows: Some OS Design Differences

 

Matthew Isaacs

Berea College

 

Abstract

This presentation provides an overview of some of the differences between the Linux and Windows operating systems.  The overall OS designs, process management systems, and memory management systems will be discussed.

 

Undergraduate student

 

 

2:50

 

How to find consensus in a family of DNA sequences?

 

Joshua W. Gilkerson, Department of Computer Science,

Jerzy W. Jaromczyk, faculty advisor

University of Kentucky

 

Abstract

A consensus sequence or string attempts to capture the common features of a family of strings.  Such consensus strings are useful in studying related genomic sequences where multiple copies of a gene may vary between individuals or species and in analyzing non-coding repeated sequences within the same genome.

 

A Steiner consensus string is a common formalization of the consensus. It is a string for which the sum of the edit distances - one of the measures for similarity of sequences - to each of the members in the family is minimized. Constructing a consensus sequence is not a computationally easy problem and no known methods are reasonable for large instances. In fact, finding an exact consensus sequence has been shown to be NP-hard(1). Therefore approximation techniques are often used to acquire meaningful results much more quickly that with exact methods. In this presentation we will show a variety of algorithms for constructing approximate Steiner consensus strings. Each of these methods may be applied to distances

different than the edit distance and they will be illustrated on families of biological sequences.

 

(1) Gusfield, Algorithms on strings, trees and sequences. 1997. Cambridge University Press.

 

Graduate Student

3:10

 

3-D Articular Cartilage Visualization from MR Images Using Expectation Maximization

 

Paul J. Thacker, Department of Computer Science,

Jerzy W. Jaromczyk, faculty advisor

University of Kentucky

 

Abstract

When following the progression of osteoarthritis, it is important to precisely determine the presence or absence of cartilage in joints such as the knee. Direct visualization from MR Images - each image representing a slice of the knee - is difficult as those images are hard to segment, i.e., to select which areas are cartilage.

 

With segmentation results from several raters and with a statistical technique known as expectation maximization it is possible to find the probability that each pixel should have been selected, thus likely belonging to the area of cartilage.

 

This process can be repeated over many image slices, with a probability attached to each voxel (3D pixel). Having gained this information, we can then visualize the most likely cartilage location. In our OpenGL implementation, we can draw a cube for each voxel above a certain probability threshold, and scale the image to have appropriate dimensions. The image can then be rotated interactively for viewing from all angles.  3D rendering gives a view of the overall cartilage structure.

 

This presentation uses results from a joint with P. Hardy and J. W. Jaromczyk.

 

Graduate Student

 

 

 

 

 

3:30

 

Diffusion Tensor Magnetic Resonance Imaging and White-Matter Fiber Tracking

 

Ning Kang , Hao Ji & Ning Cao*

Laboratory for High Performance Scientific Computing and Computer

Simulation, University of Kentucky

 

Abstract

Diffusion Tensor Magnetic Resonance Imaging (DT-MRI) is used to measure water diffusion in human body. Diffusion is the random movement of molecules. Water diffusion is anisotropic in some parts of the human body, such as brain, heart, etc. It is known that different parts of the brain are connected by `fibers'. DT-MRI can be used to study these fibers in human brain. Ning Kang in our group developed an equation based fiber tracking method to overcome the some shortcomings of current used streamline

method.

 

Graduate students

 

 

 

 

 

 

4:00

 

Support Vector Machine Approach for Partner Selection of Virtual Enterprises
 

Jie Wang* & Jun Zhang,

Laboratory for High Performance Scientific Computing and Computer

Simulation, University of Kentucky

 

Abstract

With the rapidly increasing competitiveness in global market,dynamic alliances and virtual enterprises are becoming essential components of the economy in order to meet the market requirements for quality, responsiveness, and customer satisfaction. Partner selection is a key stage in the formation of a successful virtual enterprise. The process can be considered as a multi-class classification problem. In this paper, The Support Vector Machine (SVM) technique is proposed to perform automated ranking of potential partners. Experimental results indicate that desirable outcome can be obtained by using the SVM method in partner selections. In comparison with other methods in the literatures, the SVM-based method is advantageous in terms of generalization performance and the fitness accuracy with a limited number of training datasets.

 

Graduate student

 

 

 

4:20

 

DNA computing and how it works

 

Joseph Ibershoff, Department of Computer Science,

Jerzy W. Jaromczyk, faculty advisor

University of Kentucky

 

Abstract

DNA computing is a relatively new area of computer science that performs computation in a very unique way: rather than using electric circuits and logic gates, it uses DNA molecules and chemical processes.

 

The general principle behind typical DNA computing is to use the intrinsic parallelism present in chemical reactions. The basic approach is to start with a mixture of DNA strands that generate a so-called library of all potential solutions to the problem.  Then, various operations are performed which eventually select the correct solution from the library. Typical operations which will be discussed and whose use will be illustrated include the "begins-with" selector, the "ends-with" selector, the "contains substring" selector, and "sort-by-length" (gel electrophoresis).  All of them use common biochemical techniques.

 

The purpose of this presentation is to explain the general advantages, disadvantages, and limitations of DNA computing, and to explain a typical way DNA computing is performed.  We will illustrate the basic

concepts using an application of DNA computing to the Hamiltonian path problem, similar to the first DNA computation performed by Adleman in 1993. Finally, we will briefly discuss DNA computing on microreactors and show our simulation results.

 

Graduate student

 

 

 

 

 

 

4:40

 

Mobile Agent System

 

Dianwei Han,

Laboratory for High Performance Scientific Computing and Computer

Simulation, University of Kentucky

 

Abstract

We designed a Mobile Agent System that implemented communication while running on remote machine. At first, we introduced the background of Mobile Agent systems and gave the reasons why the research is necessary. In addition, we presented how to implement some of important functions: Naming, communications, and Moving ects. Finally, we give some of available applications since we just implemented the middle-ware.

 

 

 

5:15

 

Awards and closing remarks

 

Jaleh Rezaie, Chair,

Department of Computer Science, Eastern Kentucky University