Mitigating natural disasters is one of the world’s greatest challenges, and some of the most devastating natural disasters are wildfires. To help us understand wildfires, NASA provides satellite data that measures the fire’s intensity, using the brightness of the fires as a proxy. In this code pattern, we’ll use Watson™ Studio and Watson Machine Learning to train a model with this data, allowing us to make wildfire intensity predictions using the location on a map.
The past decade has been one of the worst periods for natural disasters, and wildfires can be one of the most destructive. Call for Code is asking developers to create software solutions that help reduce wildfire risk, determine the best methods for prevention and enforcement, and aid first-responders in evacuation and firefighting operations.
For this code pattern, we use wildfire data from NASA to predict the intensity of wildfires, using Watson Studio and Watson Machine Learning. NASA provides data for various things, from weather and climate to solar flares and wildfires. This data is paid for by U.S. taxpayers and is free to use. The missing component is machine learning, which can take data and train a model to predict one of the features of the data set. For this example, we’ll grab wildfire data and build a model that can predict intensity of the fire based on latitude and longitude.
When you have completed this code pattern, you’ll understand how to:
- Use Watson Studio Machine Learning to train a model
- Gather data from NASA for wildfires
- Create a predictor for wildfire intensity based on latitude and longitude
- Add data assets and services using Watson Studio.
- Create the machine learning model in Watson Machine Learning.
- Interact with Web UI to choose the location of a fire.
- The Web UI interacts with ML model to predict the brightness of fire.
Ready to put this code pattern to use? Complete details on how to get started running and using this application are in the README.