Inspiration

It is important for the well being and financial security of farmers to have access to meaningful yield predictions. We wanted to help farmers by leveraging our technical skills.

What it does

Given a dataset of EVI values, it predicts the crop yield for a given year.

How we built it

We trained a deep neural network on EVI data.

Challenges we ran into

Data pre-processing, padding, and anomaly detection was challenging.

Accomplishments that we're proud of

Managing to create a model that predicts reasonably well. We get an error of ~11% on our validation set.

What we learned

The entire data processing pipeline from ideation, visualisation, pre-processing, and modelling.

What's next for CNN_Cropper

Exploring other methods of modelling such as Gaussian Processes, ensemble models, and time series analysis.

Built With

Share this project:

Updates