Research Symposium

22nd annual Undergraduate Research Symposium

Jennifer Pierre Her/She Poster Session 1: 9:00 - 9:45/Poster #28


IMG_0293.jpg

BIO


My name is Jennifer Pierre, and I am from Naples, Florida. Iā€™m interested in researching historical events, social sciences, and natural sciences. Currently, I am researching a subject related to neural networks associated with hurricanes. My career goals are to find a job where I can travel while working for a good company.

Prediction of Hurricane Paths through Neural Networks

Authors: Jennifer Pierre, Diogo Nunes De Oliveira
Student Major: Computer Science
Mentor: Diogo Nunes De Oliveira
Mentor's Department: Information Technology
Mentor's College: College of Communication & Information at Florida State University
Co-Presenters: Natalia Chamizo, Ryan Fontaine

Abstract


Background and purpose: Studies suggest that there are alternative methods to precisely predict hurricane trajectories. We examined past hurricanes to determine a pattern and cross-reference these neural-networks to create a new system of predicting through coding via python.ā€‹

Methods: Information was gathered from the past 20 years in the panhandle area of Florida. We observed and analyzed hurricane hits and considered wind patterns on the date and location of the event. Data was also gathered from each big hurricane that happened in the area specifically from the past 10 years such as Hurricane Michael and Irma. Information being collected is later converted into data points to develop a pattern including the date of landfall, the strength of the hurricane, diameter, etc.ā€‹

Results: Not yet determined (still in progress).ā€‹

Conclusions: In progress. Idea is that when peers are done implementing perception in regard to the hurricane trajectories through the python coding system, we can try different algorithms and begin to input data points from past hurricane events.

Keywords: Hurricanes, Trajectory, Perceptron