MCDFN: Supply Chain Demand Forecasting via an Explainable Multi-Channel Data Fusion Network Model Integrating CNN, LSTM, and GRU

2024-05-24

arXiv.org

AI Summary

To address the challenges of modeling seasonality and abrupt demand fluctuations in supply chain forecasting, as well as the limited interpretability of deep learning models, this paper proposes an interpretable multi-channel fusion network. The architecture parallelly integrates CNN (for local temporal pattern extraction), LSTM, and GRU (for long-term dependency modeling) to jointly learn spatiotemporal features. We innovatively incorporate ShapTime and permutation-based feature importance to enable fine-grained attributional interpretation. To our knowledge, this is the first work to systematically integrate and empirically validate the synergistic benefits of these three major RNN variants. On standard benchmark datasets, 10-fold cross-validation yields MSE=23.57, RMSE=4.86, MAE=4.00, MAPE=20.16%, and Theil’s U=0.118; a t-test (α=0.05) confirms no statistically significant prediction bias. The model achieves both high predictive accuracy and decision transparency.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Machine Learning: Deep Neural Architectures and Foundation Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Accurate demand forecasting is crucial for optimizing supply chain management. Traditional methods often fail to capture complex patterns from seasonal variability and special events. Despite advancements in deep learning, interpretable forecasting models remain a challenge. To address this, we introduce the Multi-Channel Data Fusion Network (MCDFN), a hybrid architecture that integrates Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM), and Gated Recurrent Units (GRU) to enhance predictive performance by extracting spatial and temporal features from time series data. Our comparative benchmarking demonstrates that MCDFN outperforms seven other deep-learning models, achieving superior metrics: MSE (23.5738), RMSE (4.8553), MAE (3.9991), and MAPE (20.1575%). Theil's U statistic of 0.1181 (U<1) of MCDFN indicates its superiority over the naive forecasting approach, and a 10-fold cross-validated statistical paired t-test with a p-value of 5% indicated no significant difference between MCDFN's predictions and actual values. We apply explainable AI techniques like ShapTime and Permutation Feature Importance to enhance interpretability. This research advances demand forecasting methodologies and offers practical guidelines for integrating MCDFN into supply chain systems, highlighting future research directions for scalability and user-friendly deployment.

Problem

• Enhance supply chain demand forecasting accuracy
• Address interpretability in deep learning forecasting models
• Integrate spatial and temporal features for better predictions

Innovation

• MCDFN integrates CNN, LSTM, and GRU networks
• Uses ShapTime and Permutation Feature Importance
• Outperforms seven deep-learning models in benchmarks