Multi-Scale Food Flow Prediction using Graph Neural Networks

A project leveraging Graph Neural Networks (GNNs) to predict food flows between counties and FAF zones for economic planning, infrastructure development, and policy-making. This model predicts food trade flows between U.S. counties and Freight Analysis Framework (FAF) zones using Graph Neural Networks (GNNs). It addresses the challenges of sparsity in trade data by applying a two-stage hurdle model that distinguishes between the presence and magnitude of trade.

Great News! The portal is now online at https://gnnfoodflowportal.pods.icicleai.tapis.io/

Please go to https://github.com/ICICLE-ai/GNNFoodFlowPortal/ for more updated information and better accessability

Model Details

• Developed by: Qianheng Zhang & ICICLE Team
• Funded by: NSF AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE) (OAC 2112606)
• Model type: Graph Neural Network (GAT and GCN variants)
• Language(s): English (for documentation and metadata)
• License: MIT License
• Framework: PyTorch, PyTorch Geometric

References

Data Sources

• Trade Data: FAF5.6.1 SCTG1 commodity flow data (code/data/FAF5_SCTG1.csv)
• Geographic Information:
  • County shapefiles (code/data/shapefiles/cb_2017_us_county_500k/cb_2017_us_county_500k.shp) https://github.com/GeoDS/GNNFoodFlow/blob/master/portal/data/shapefiles/cb_2017_us_county_500k/cb_2017_us_county_500k.shp
  • State shapefiles (code/data/shapefiles/cb_2018_us_state_20m/cb_2018_us_state_20m.shp) https://github.com/GeoDS/GNNFoodFlow/blob/master/portal/data/shapefiles/cb_2017_us_county_500k/cb_2018_us_state_20m/cb_2018_us_state_20m.shp
  • FAF zones shapefiles (code/data/shapefiles/2017_CFS_Metro_Areas_with_FAF/2017_CFS_Metro_Areas_with_FAF.shp) https://github.com/GeoDS/GNNFoodFlow/blob/master/portal/data/shapefiles/2017_CFS_Metro_Areas_with_FAF/2017_CFS_Metro_Areas_with_FAF.shp
• FAF Economic Indicators: FAF-level economic data (code/data/faf_features_aligned_filtered.csv)https://github.com/GeoDS/GNNFoodFlow/blob/master/code/data/faf_features_aligned_filtered.csv
• County Economic Indicators: County-level economic data (code/data/county_aligned_filtered.csv) https://github.com/GeoDS/GNNFoodFlow/blob/master/code/data/county_aligned_filtered.csv
• Distance Information: FAF Distance Matrix (code/data/FAF_Distance_Matrix.csv) https://github.com/GeoDS/GNNFoodFlow/blob/master/code/data/FAF_distance_matrix.csv

Acknowledgements

National Science Foundation (NSF) funded AI institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE) (OAC 2112606)

Future Work

• Extending the model to handle inter-county food trade flow predictions
• Refining the model to capture more granular food trade patterns