team photo

Team 13

Team Members

Faculty Advisor

Benjamin Goh
Mazin Alam
Alexander Ercolani
Michael Picard
Obumneme Nkwo

Wei Wei


UConn Computer Science & Engineering Department

sponsored by
Sponsor Image Not Available

Wildfires are a destructive force that can be created by humans and nature. They are challenging to predict, although they are most likely to occur in areas that are very dry. For this project, we are going to focus on wildfires in California because they are common in that area, and due to climate change, they are occurring more frequently and are spreading over larger areas. This project, Trackit!, is a web-based wildfire predictor application. It allows users to interact with a machine learning model that predicts the likelihood of wildfires in California. Users will click on a location marker pin on our interactive map, and a pop up will alert the user of the risk level of a wildfire in the selected area in the next two weeks. They can use this to help ensure their safety when traveling to or within California. We trained a deep neural network using historical wildfire data its associated weather data to predict the probability of a wildfire occurring. The weather data was web scraped from the Farmer's Almanac website, which provides accurate historical data. The wildfire data was exported from the National Oceanic and Atmospheric Administration (NOAA) website.