Ashley Abare
Advisor: Ed Harcourt, Computer Science
Geographic Information System: Weather Data Management and Manipulation
In the pursuit of efficiency, I have created a useful Geographic Information System database populated automatically from three local weather stations located at: Saint Lawrence's Ecologically Sustainable Landscape (ESL), Little River Nature Area, and Waterman Hill. The database contains climate and atmospheric information from each station which vary in vegetation and elevation. Administrators and everyday users will benefit from this new database since it will improve the visibility and increase the accessibility of data through the use of tables which can be merged, updated, and archived much better than the previous primitive system which had to be updated manually. |
Andrea Adams
Advisor: Ed Harcourt, Computer Science
A Relational Database for the c.elegans nematode
Bioinformatics is a field in which computer systems are used to process experimental data from biological experiments. Working with members of the Biology department, I have constructed a relational database system to aid them in processing the data from an experiment involving the results on genes expressed in neurons and mitochondria in the nemetode c.elegans when it is raised in an environment devoid of oxygen. To accomplish this, I mined data from the existing SAGE and Mitores databases to construct two relational databases: one detailing all of the genes expressed in neurons, and the other detailing the genes expressed in mitochondria. Then, once experimental data has been loaded into the database, the biologists use a web interface to access and query the databases. |
|
Nick Alena
Advisor: Robin Lock, Statistics
Applications of Linear Discriminant Analysis
Is it possible to use numerical measurements of an insect to distinguish its species? Through the examination of Midge data, we will see that this is possible with the statistical method, Linear discriminant analysis. LDA is a statistical method used to identify group membership using a linear combination of features. The techniques extend concepts from ANOVA (analysis of variance) and regression analysis. We examine methods for choosing good discriminanting variables and producing a linear combinition that best distinguishes between the groups. |
|
Allison Avrich and James Padgett
Advisor: Ed Harcourt, Computer Science
Pathfinding using Lego NXT Robot with NXC
This project explores algorithmic and heuristic techniques of pathfinding by programming a Lego Mindstorms NXT Robot using the Not Exactly C(NXC) programming language. The robot itself is a three-wheeled(two powered, one support wheel) vehicle with a light and color sensor adjacent to each other attached to the front of the robot. By using a black tape track on a white background, the robot is programmed it interpret the input of its sensors to detect the track and follow it. A variety of different tracks were used to test the capabilities of the robot. Each track was designed to be more difficult to navigate than the previous track. Topics within the scope of this project include: different algorithmic and heuristic search, pathfinding, and path-following techniques; multitasking and concurrent programming; and advanced levels of complexity, such as obstruction avoidance. |
| |
|
Nicholas Bonsack
Advisor: Ed Harcourt, Computer Science
C++ Interface for Graph Structures
My project is a C++ interface for graph structures that can be used by software developers to visualize graphs. For software developers this interface inherits a graph class, where it can be given instructions to print out the contents of a graph in a particular manner. This allows them to see the graph as an image, which is easier to interpret than a simple text dump on a terminal screen. The interface is made to work with simple graph structures at present, but the design is extensible enough that with some tweaking, it can work with any graph structure provided that an iterator for it exists. When called upon to draw a graph, this interface takes two passes through the graph - one to collect its contents, a second to determine its paths - and draws out the vertices and edges to an external structure. To accomplish its goals, the interface relies on templated classes and multiple class inheritance. Both of these are techniques that can only be presently done under the C++ language. |
|
Benjamin Case and Stephen Ciesla
Advisor: Ed Harcourt, Computer Science
Beowulf Cluster Computing
Scientific computing is requiring greater and greater processing power. Cluster computing provides a inexpensive means of creating powerful parallel processing computers. We construct a Beowulf Linux cluster for Bioinformatics work and show the research and construction of the cluster as well as the mechanics of how parallel processing systems are used to tackle large scale problems. |
Domingo Garcia
Advisor: Dr. Knickerbocker, Computer Science
Media Center
The purpose of this project is to create a new efficient way to view, browse, and access personal media including RSS feeds. Using software engineering principles, the incremental model was used for the development of this project. Java Swing, a programming language that is used to create graphical user interfaces, was used to create a unique thumbnail view of all media, and navigation through data with just a few clicks. |
Alan Marcero
Advisor: Ed Harcourt, Computer Science
Web Development Using the Model-View-Controller Programming Architecture
The beginning of the end for PHP? As the first "web 2.0" framework released, Ruby On Rails (often shortened to Rails), encourages and includes constructs for development using techniques such as: agile methods, don't repeat yourself (DRY) coding, model-view-controller architecture, and convention over configuration. All of which combine to create a framework that allows the the development of web applications to be faster than ever before. Applications that would take months in PHP, take only weeks in Rails. My project, available at http://www.alanmarcero.com, is a use of this framework on a large-scale, fully database centric web application. The application features several "web 2.0" traits such as: strict validating XHTML/CSS, multiple Ajax implementations, animated Javascript effects, a fully functional blog engine, a dynamic RSS feed, and a dynamic "live" search. |
Yordan Minev
Advisor: Michael Schuckers, Statistics
ROC Confidence Regions Using Radial Sweep
A biometric authentication system matches physiological characteristics to a database of such characteristics. In biometric authentication, genuine users are generally those that the system should accept and imposters are those that the system should reject. One methodology for evaluating the matching performance of biometric authentication systems is the receiver operating characteristics (ROC) curve. The ROC curve graphically illustrates the relationship between type I and type II statistical classification errors when varying a threshold across a genuine and an imposter match score distributions. The performance of each biometric system can be estimated via a confidence region for a ROC curve of that system's performance. In this project ROC confidence regions will be created using radial sweep method. Radial sweep is based on converting the type I and type II errors to polar coordinates. The technique of bootstrapping will be utilized to estimate the variability of each point on an individual ROC curve. Simulations will be performed using real biometric match score data. A radial sweep method for comparing two ROC curves will be discussed. |
|
Julia Palmateer
Advisor: Robin Lock, Statistics
Assessing Ratings Methods for College Hockey Teams
Ranking college sports teams that play in leagues of different strengths can be challenging. A number of methods have been proposed to calibrate teams while accounting for performance and strength of schedule. We use Monte Carlo simulations of hundreds of seasons to compare several existing models, such as raw winning percentage, rating percentage index (RPI), Bradley-Terry (KRACH) and Poisson scoring rates (CHODR), for rating college ice hockey teams. By using these simulations we can examine common questions such as what are the pros and cons of being a strong team in a weak league versus being an average team in a strong league. |