Home
Softono
SupplyGraph

SupplyGraph

Open source Jupyter Notebook
GitHub
96
Stars
20
Forks
0
Issues
0
Watchers
1 year
Last Commit
ML Frameworks Data Visualisation

About SupplyGraph

SupplyGraph is a comprehensive benchmark dataset designed to advance the application of Graph Neural Networks in supply chain planning. Recognizing that supply networks are inherently graph-structured, this resource addresses the critical lack of real-world data and conceptual frameworks for GNN-driven supply chain solutions. Developed from data provided by a leading Fast-Moving Consumer Goods company in Bangladesh, the dataset supports research into optimizing complex problems such as production planning, demand forecasting, and anomaly detection. The project evaluates state-of-the-art GNN models across six distinct analytics tasks on both homogeneous and heterogeneous graph structures. Experimental results demonstrate that GNN-based approaches significantly outperform traditional statistical machine learning and standard deep learning models. Specifically, these graph methods achieve performance gains of 10 to 30 percent in regression and classification tasks, with anomaly detection improvements ranging bet

Platforms

Web Self-hosted

Languages

Jupyter Notebook

Links

Source Code
c
Published by
ciol-researchlab
Visit View Profile

README.md

View on GitHub

SupplyGraph: A Benchmark Dataset for Supply Chain Planning using Graph Neural Networks

arxiv Kaggle

đź“Ś TL;DR: This paper introduces a real-world graph dataset empowering researchers to leverage GNNs for supply chain problem-solving, enhancing production planning capabilities, with benchmark scores on six homogeneous graph tasks.

Abstract: Graph Neural Networks (GNNs) have recently gained traction in transportation, bioinformatics, language and image processing, but research on their application to supply chain management remains limited. Supply chains are inherently graph-like, making them ideal for GNN methodologies, which can optimize and solve complex problems. The barriers include a lack of proper conceptual foundations, familiarity with graph applications in SCM, and real-world benchmark datasets for GNN-based supply chain research. To address this, we present a multi-perspective real-world benchmark dataset from a leading FMCG company in Bangladesh, focusing on supply chain planning. We discuss various supply chain tasks using GNNs and benchmark several state-of-the-art models on homogeneous and heterogeneous graphs across six supply chain analytics tasks using this dataset. Our analysis shows that GNN-based models consistently outperform statistical Machine Learning and other Deep Learning models by around 10-30% in regression, 10-30% in classification and detection tasks, and 15-40% in anomaly detection tasks on designated metrics. With this work, we lay the groundwork for solving supply chain problems using GNNs, supported by conceptual discussions, methodological insights, and a comprehensive dataset.

For coding instructions, check code\coding_directions.md

Citation:

@misc{wasi2024supplygraph,
      title={SupplyGraph: A Benchmark Dataset for Supply Chain Planning using Graph Neural Networks}, 
      author={Azmine Toushik Wasi and MD Shafikul Islam and Adipto Raihan Akib},
      year={2024},
      eprint={2401.15299},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}