Beyond Accuracy: Evaluating Forecasting Models by Multi-Echelon Inventory Cost

TL;DR

This paper presents a comprehensive evaluation of various forecasting models integrated with inventory optimization, demonstrating that advanced models like Temporal CNN and LSTM can significantly reduce costs and improve service levels in supply chain systems.

Contribution

It introduces a unified digital pipeline combining forecasting and inventory optimization, and evaluates model performance on multi-echelon supply chain scenarios using real-world data.

Findings

Temporal CNN and LSTM models outperform statistical baselines in cost reduction.

The integrated pipeline is robust across different supply chain configurations.

Advanced models improve fill rates and scalability in inventory management.

Abstract

This study develops a digitalized forecasting-inventory optimization pipeline integrating traditional forecasting models, machine learning regressors, and deep sequence models within a unified inventory simulation framework. Using the M5 Walmart dataset, we evaluate seven forecasting approaches and assess their operational impact under single- and two-echelon newsvendor systems. Results indicate that Temporal CNN and LSTM models significantly reduce inventory costs and improve fill rates compared to statistical baselines. Sensitivity and multi-echelon analyses demonstrate robustness and scalability, offering a data-driven decision-support tool for modern supply chains.