Math 627 - Introduction to Parallel Computing
Spring 2011 - Matthias K. Gobbert
Presentations of the Class Projects

Friday, May 20, 2011, 02:00 p.m., SOND 207

  1. 01:30-01:45
    A Parallel Simulation of the Evolution of Transcription Factor Binding Sites
    Robert Forder, Department of Mathematics and Statistics
    The analysis of transcription factor binding motifs may aid in understanding the process by which transcription factors recognize their binding sites. We wish to investigate the likelihood that transcription factors use correlations between positions in potential binding sites as a critera for recognition. We implement a genetic algorithm in parallel using a simple server-client organization to simulate the evolution of these binding sites. We then evaluate the performance of this application and conclude that exhibits excellent speedup and efficiency. This work is collaborative with Joseph Cornish and Dr. Ivan Erill from the Department of Biologicial Sciences and with Dr. Matthias Gobbert from the Department of Mathematics and Statistics as part of the project Interdisciplinary Training for Undergraduates in Biological and Mathematical Sciences (UBM).

  2. 01:50-02:05
    A Simulation Study of Parameter Estimation Methods for the Dirichlet-Multinomial Distribution using R in a Parallel Framework
    Amanda Peterson, Department of Mathematics and Statistics
    In the 1998 paper entitled "Large Cluster Results for Two Parametric Multinomial Extra Variation Models," Nagaraj K. Neerchal and Jorge G. Morel developed an approximation to the Fisher Information matrix used in the Fisher Scoring algorithm for estimating the parameters of the Dirichlet-multinomial distribution. They performed simulation studies, comparing the results of the approximation to the results of the usual Fisher Scoring algorithm, for varying dimensions of the parameter vector. In this study, parallel computing in R is utilized to extend the previous simulation studies to larger dimensions. Additionally, the Fisher Scoring algorithm and the Maximum Likelihood Estimation method are compared. This work is joint with Dr. Nagaraj Neerchal and Andrew Raim.

  3. 02:10-02:25
    Modeling Long Term Flow and Contaminate Transport in Two-Dimensional Geometries with Semi-Impervious Heterogeneities as Related to the Delmarva Peninsula
    Mattie Whitemore, Department of Mathematics and Statistics
    Flow and transport of nitrate contamination from industrial farms in groundwater is studied to determine long term changes to water quality in the Delmarva Peninsula. In general, groundwater residence times decrease as one moves closer to a stream. However, due to the heterogeneous nature of the aquifer in our geographical region this is not always the case. A two-dimensional model is constructed in COMSOL Multiphysics to illustrate the changes to river quality due to six different cases for semi-impervious clay banks in the cross-sectional geometry, and compared to one geometry free of heterogeneities. Two coupled PDEs are prescribed to the geometry to quantify flow and contamination transport in the system. The model accurately demonstrates the known groundwater age in the Delmarva Peninsula. The study is carried out for 150 years in order to examine how heterogeneities in the aquifer change the residence time for areas furthest away from a river, as well as changes in the total concentration of nitrogen reaching a river. This work is collaborative with Dr. Baker of the UMBC Geography and Environmental Science Department and Dr. Peercy of the UMBC Mathematics and Statistics Department.

Copyright © 2001-2011 by Matthias K. Gobbert. All Rights Reserved.
This page version 1.0, May 2011.