Skip to content

Harness the power of machine learning to forecast rice and wheat crop yields per acre in India, aiming to empower smallholder farmers, combat poverty and malnutrition, utilizing data from Digital Green surveys to revolutionize agriculture and promote sustainable practices in the face of climate change for enhanced global food security.

License

Notifications You must be signed in to change notification settings

association-rosia/crop-yield-estimate

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🌾 Crop Yield Estimate

This challenge focuses on leveraging machine learning to predict rice and wheat crop yields per acre in India, with the aim of empowering smallholder farmers and addressing issues of poverty and malnutrition. The data collected by Digital Green through surveys provides insights into farming practices, environmental conditions, and crop yields. The ultimate goal is to revolutionize Indian agriculture, offer a global model for smallholder farmers, and contribute to sustainable farming practices amid climate change, thereby advancing global food security.

This project was made possible by our compute partners 2CRSi and NVIDIA.

🏆 Challenge ranking

The score of the challenge was the RMSE.
Our solution was the best one (out of 678 teams) with a RMSE equal to 100.3610312 🎉.

The podium:
🥇 RosIA - 100.3610312
🥈 ihar - 100.6819477
🥉 belkasanek - 102.4325999

🛠️ Data processing

Pre-processing

GReaT

We used GReaT LLM (GPT-2) implementation to generate new observations and impute missing values.
Here is how this proposed method works to create and impute data:

🏛️ Model architecture

#️⃣ Command lines

Launch a training

python src/models/train_model.py --estimator_name <estimator_name> --task <task> --nb_agents <nb_agents>

View project's runs on WandB.

Create a submission

python src/models/predict_model.py --ensemble_strategy <ensemble_strategy> --class_id <class_id_1> <class_id_2> <class_id_3> --low_id <low_id_1> <low_id_2> <low_id_3> --medium_id <medium_id_1> <medium_id_2> <medium_id_3> --high_id <high_id_1> <high_id_2> <high_id_3>

🔬 References

Hwang, Y., & Song, J. (2023). Recent deep learning methods for tabular data. Communications for Statistical Applications and Methods, 30(2), 215-226.

Shwartz-Ziv, R., & Armon, A. (2022). Tabular data: Deep learning is not all you need. Information Fusion, 81, 84-90.

Kavita, M., & Mathur, P. (2020, October). Crop yield estimation in India using machine learning. In 2020 IEEE 5th International Conference on Computing Communication and Automation (ICCCA) (pp. 220-224). IEEE.

Borisov, V., Seßler, K., Leemann, T., Pawelczyk, M., & Kasneci, G. (2022). Language models are realistic tabular data generators. arXiv preprint arXiv:2210.06280.

📝 Citing

@misc{UrgellReberga:2023,
  Author = {Baptiste Urgell and Louis Reberga},
  Title = {Crop Yield Estimate},
  Year = {2023},
  Publisher = {GitHub},
  Journal = {GitHub repository},
  Howpublished = {\url{https://github.com/association-rosia/crop-yield-estimate}}
}

🛡️ License

Project is distributed under MIT License

👨🏻‍💻 Contributors

Louis REBERGA

Baptiste URGELL

About

Harness the power of machine learning to forecast rice and wheat crop yields per acre in India, aiming to empower smallholder farmers, combat poverty and malnutrition, utilizing data from Digital Green surveys to revolutionize agriculture and promote sustainable practices in the face of climate change for enhanced global food security.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published