Yield Predictor

Inspiration

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for Yield Predictor

Our goal was to Predict the Future - 60 and 90- day forecast yield simulation. We approached this by aiming to build a predictive model that estimates expected yields for the current growing season based on near-term weather forecasts.

Our two main datasets were the daily weather observations and the yearly yield data. We cleaned the data to be between 2010 and 2024, and averaged the weather between the growing season months (April-September), then joined them with a third dataset, converting longitude and latitude to U.S. state names. After uploading this data into a Databricks workspace, we trained a model on this data and had it predict the yield amount based on the weather.

Using the open source library, Streamlit, we created an app that could display this data once we connected the databricks workspace in the back end.

Then with the website, we connected it again to databricks in order for it to be hosted non locally