Forest Fire Prediction

We knew that wildfires were a big problem in the last couple of years in California, and we wanted to try and find a way to limit the fires from spreading. Our project predicts the amount of land the next wildfire burns, which can help limit the spread and also help people evacuate quicker. We used a data set from kaggle.com and fed it into a boosting algorithm. We used grid-search to improve upon the hyperparameters to select the best model. The major challenge was that the dataset was pretty small, so it was hard to create an accurate model. Later, we could expand this into an app or a website which could help firefighters save lives.