Inspiration
We saw this as a great opportunity to learn and flex some of our skills.
What it does
We built an XGBoost model to predict the severity of wildfires in California using historic wildfire and weather data. We focused on California wildfires that occurred between January 1, 2000 through December 31, 2015 in the following counties: Riverside, Los Angeles, El Dorado, San Bernardino, and San Diego. Daily measures of average precipitation, temperature, and wind speed were coupled with the wildfire data.
How we built it
We built it using a Kaggle wildfire dataset, downloading various county-level weather datasets from NOAA, and collaborated using Google Colab before moving everything over into SageMaker Studio Lab.
Challenges
Some of the challenges we faced included: finding a reliable dataset, cleaning the dataset sufficiently, deciding on a methodology for analysis, and tinkering with final methodology to improve results
Accomplishments
We have never competed in a Hackathon before and we were proud to participate this year!
Log in or sign up for Devpost to join the conversation.