article · Frontiers in Big Data
Accurate retail sales forecasting demands models that deliver high predictive accuracy while scaling efficiently across complex, large-scale data. A hybrid framework named XGB-ANN-Attn addresses this challenge by combining feature-level representations. The architecture fuses leaf embeddings generated by extreme gradient boosting with deep neural representations, enabling joint modelling of low-order statistical dependencies and high-order non-linear feature interactions. A lightweight attention mechanism dynamically calculates instance-specific feature importance, keeping computational complexity linear relative to feature dimensionality. When evaluated on Walmart data, the framework achieved a score of 0.9568. The design supports parallelisation and distributed deployment, and features strong alignment between attention weights and SHAP explanations. This creates a scalable and transparent forecasting method suitable for enterprise retail decision support systems.
Retail businesses often struggle to balance complex machine learning accuracy with model interpretability and computational speed. By merging tree-based models with neural networks and attention mechanisms, this approach enables companies to forecast sales accurately without creating opaque black-box systems. It allows decision-makers to understand the underlying drivers behind demand predictions while handling massive data streams efficiently.
The framework is suited for large-scale retail analytics, inventory planning, and automated commercial decision-support systems. Potential users include enterprise retailers and supply chain managers managing complex sales data. Having been applied and tested on Walmart retail data with distributed deployment and parallelisation capabilities, the technology appears positioned as an applied framework ready for integration into enterprise big-data pipelines.
AI-generated from the published abstract. Always read the original work before citing.
Introduction: Accurate retail sales forecasting in modern data-intensive environments requires models that not only achieve high predictive accuracy but also scale efficiently as data volume and feature complexity increase. This study proposes a novel cognitively inspired hybrid framework, XGB-ANN-Attn, for interpretable and scalable retail analytics. Methods: The model introduces feature-level fusion by integrating XGBoost-derived leaf embeddings with deep neural representations, enabling joint modeling of low-order statistical dependencies and high-order nonlinear feature interactions. A lightweight attention mechanism dynamically assigns instance-specific feature importance, enhancing both predictive performance and interpretability while maintaining linear computational complexity with respect to feature dimensionality. Unlike conventional ensemble approaches that operate at the decision level, the proposed framework integrates representations at the level of representations, reducing redundancy and improving computational efficiency. The model is designed for scalability, combining the log-linear complexity of gradient boosting with the linear scaling properties of neural networks and attention mechanisms. Results: = 0.9568 on Walmart. Furthermore, the framework supports parallelization and distributed deployment, making it suitable for large-scale retail systems. Discussion: The alignment between attention weights and SHAP explanations provides transparent and actionable insights. The results confirm that the proposed approach offers a scalable, interpretable, and high-performance solution for AI-driven decision support systems in big-data retail environments.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.3389/fdata.2026.1889044
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
