omniscient corn and beans

• 

  Interface of our application

• 

  Sample of rows from our dataset

• 

  Graph of predicted yield vs maximum temperatures

• 

  Graph of predicted yield vs precipitation

Inspiration

For our project we created, trained, and tested a gradient boosting predictive model to estimate the expected yield for the current growing season based on weather data from across the nation. One of the major challenges we faced was training the model to accurately predict the crop yield. We had to play around with multiple different hyperparameters using grid search to find optimal values and sample data sizes. We experimented with different models and dataset sample sizes to obtain an acceptable test loss value while validating with RSME and MAE values. We then embedded our model into our React application to provide a readable and accessible interface that users can input expected weather conditions into to learn their predicted yield.

Built With

• express.js
• javascript
• python
• react
• sql