Inspiration
I was born in New Orleans and I know the impact hurricanes can have on peoples lives. I was thinking, if we can better predict the behaviors of mother nature, we can save a lot of time, money and lives. Although better predictive models exist, I wanted to try my hand at this problem by applying some of the concepts I learned at UTA.
What it does
Given a hurricane's current position, max wind and classification, predict where it will be in the future.
How we built it
I built it using matlab and in about 5 hours of time (including research). It uses hurdat2 dataset and Matlab's fitlm() function to create a linear model which takes in current, position, max wind and classification to output predicted coordinates. I then plotted those coordinates using geoshow().
Challenges we ran into
Getting the dataset into matlab was harder than expected.
Accomplishments that we're proud of
I learned alot about Matlab and all of it's cool libraries .
What we learned
I learned about the existence of LTSM which is a deep learning model which would probably be better suited for this problem.
What's next for Predictive Modeling of Hurricane Tracks
I plan to try and increase model accuracy by perhaps using a different model or switching variables.


Log in or sign up for Devpost to join the conversation.