For wheat, the median estimate is zero productivity change, but some models predict large positive increases and a couple expect enormous drops. Luckily, as this experiment is based on simulated data, we have ground truth to compare to at the end of the competition. Join the competition now!

How far into the future can we make good predictions? Which types of models or training methods do better or worse? Join the challenge to help us answer these questions! And join the AgML mailing list to stay up-to-date with this and our other activities: agml.org

Sounds easy? In a practice run, using a different crop model, here's how six different ML models predicted the global annual maize prediction (black line shows the true value). Which models do you think each line corresponds to? We were surprised by the results!

Your mission is to predict (simulated!) end-of-season wheat and maize yields using the daily weather experienced during the growing season. The challenge is that you must predict yields in the future (2021 to 2100) but will only receive training data from the last few decades...

Also liked this paper (and the article in Patterns it discusses)! Super interested to read your paper when it's out.

This is my favorite image for communicating the idea of compensating errors/right answer wrong reason 😂