CropGym: a Reinforcement Learning Environment for Crop Management
Abstract
Nitrogen fertilizers have a detrimental effect on the environment, which can be reduced by optimizing fertilizer management strategies. We implement an OpenAI Gym environment where a reinforcement learning agent can learn fertilization management policies using process-based crop growth models and identify policies with reduced environmental impact. In our environment, an agent trained with the Proximal Policy Optimization algorithm is more successful at reducing environmental impacts than the other baseline agents we present.
Download full text in pdf format
Published as:
H. Overweg,
H. N. C. Berghuijs,
I. N. Athanasiadis,
CropGym: a Reinforcement Learning Environment for Crop Management,
Computing Research Repository,
AIMOCC 2021 Workshop at 9th Int'l Conf Learning Representations (ICLR 2021),
2021, doi:10.48550/arXiv.2104.04326.
You might also enjoy (View all publications)
- Better together: Evaluating the complementarity of Earth embedding models
- Cooperative Research Activities in a Cloud-Based Virtual Research Environment: From Notebooks to Workflows, Digital Twins, and AI Agents
- The role of observations in calibrating crop models: A multi-model analysis of rice phenology