Inferknow

The project we made was inspired by the fact that forest fires are one of the more underlooked natural disasters that occur frequently across the globe. Amid the recent one in LA, we decided to develop a machine learning model that can predict upcoming forest fires in regions whilst also accounting for human made reasons since 90% of forest fires are man-made. We first extracted and cleaned various datasets including recent fires, weather, population density and a few. Our project focused on California since it is the most forestfire prone state in the country. Then we trained our machine learning model. This is where most of the challanges arrived since initially we used different models but all gave results that were not very convincing. Eventually we got a model which gave us a reasonable accuracy and went with it. This was our first time creating something from scratch using machine learning models and the experience was great.

Built With

• bash
• html
• numpy
• python
• shell