| Science
Fair Projects |
|
2000 |
- 1) Joey Echeveria
- The Effects of
Node Number on the
Performance of a Beowulf Super Computer Joey used Linux operating
system to linked 24 of the lab's computers into a Beowulf system so that
they functioned as though they were a single super computer. For a period
of time Southside had the fastest high school computer system in the
world. This project received media attention from local radio TV and
newspapers.
- Category: computer science
- Awards: 1st Place Computer Science Div.,
Honorable Mention Overall, 4 special awards. Attended
the Intel International Science fair as an observer. Although Joey did not
have the chance to display his project it received media attention from
radio, television, and newspaper. His project was a major factor in the
scholarship he received from the NSA that allowed him to attend Carnegie
Mellon University, all expenses paid.
|
- 2) Greg Grothouse
- Neural Network Vision Recognition System:
Greg designed and built a digital camera/computer system which
learned to recognize numbers by using a neural network. Wrote the neural
network software. This project made it all the way to the International
Science Fair.
- Category: engineering, computer
science
- Awards: 1st Place Engineering Div., 2nd
Place Overall, 4 special awards. Attended
the Intel International Science fair.
|
|
2003 |
- 4) Aaron Cheung
- Can a Computer
Learn by Accumulating Past Experience: This project’s purpose was to
create a artificial intelligence learning algorithm that demonstrated a
computer’s learning capabilities. The model of tic-tac-toe was chosen
and the computer demonstrated amazing learning capabilities. In effect,
the computer program learned the rules and strategies of tic tac toe with
no guidance.
- Category: computer science
- Awards: 4th place
overall, 3 special awards. Attended the Intel
International Science fair as an observer.
|
- 8) Nhan Nguyen
- Training Neural
Networks with Genetic Algorithms: The project looked at different
methods of training artificial neural networks. It compared the
performance of learning under the traditional backpropagation algorithm to
a less traditional use the genetic algorithm to train neural networks.
Training under the genetic algorithm, the neural network was slightly more
accurate. However, the training process took much longer than the
backpropagation training.
- Category: computer science
- Awards: 2nd
place overall, 3 special awards. Attended the
Intel International Science fair. Won 4th place in computer science and
will, subsequently, have an asteroid named in his honor. Received a $500
award from the American Association for Artificial Intelligence and a
$40,000 scholarship offer to attend Florida Institute of Technology.
|
- 11) Jeremy
VanderKnyff - Approximating
Multi-Body Orbits Using Kinematic Newtonian Analysis and Investigating the
Effects of Velocities Above and Below Critical Velocity on an Orbiting
Body: Using Newton’s basic laws
of motion, the program accurately models the way planets in a solar system
interact with each other. Using actual values for the masses and distances
of the nine planets and the Sun in our Solar System, the program
calculates the correct orbital patterns and periods for the planets down
to the hour. This is one of the closest approximations of orbiting bodies
performed using Newtonian analysis.
- Category: physics
- Awards: 3rd place overall, 2 special
awards. Won the opportunity to attended the Intel
International Science fair as an observer but was unable to go.
|
|
|
