Back to search

Georgia Institute of Technology

Advanced Data-Driven Methodologies for Enhanced Demand Forecasting in Supply Chain Management

Abstract

dc:description.abstract

In the complex and dynamic realm of supply chain management, the accuracy of demand forecasting is paramount for achieving operational efficiency, reducing costs, and enhancing customer satisfaction. This dissertation introduces a series of innovative, data-driven methodologies that significantly enhance the precision of demand forecasts across various facets of the supply chain. By integrating cutting-edge analytics and forecasting techniques, this research addresses critical challenges encountered by logistics and retail sectors, especially in light of recent global disruptions such as the COVID-19 pandemic. Chapter 2 presents a novel dynamic forecasting method specifically designed for predicting parcel arrivals at logistics hubs. As e-commerce continues its rapid expansion, logistics hubs are under increasing pressure to manage incoming parcel volumes efficiently. This chapter elucidates the development and implementation of an ensemble forecasting model that leverages both historical data and real-time parcel tracking information to predict short-term parcel arrival rates. This innovative approach enables logistics operators to optimize resource allocation, enhance throughput, and mitigate operational uncertainties. Chapter 3, expands the scope of retail demand forecasting by incorporating a spatial-temporal analytical framework. Recognizing that both geographic and temporal factors significantly influence demand, this chapter unveils a clustering-based ensemble model for forecasting. It integrates augmented fuzzy c-means clustering with advanced time series and multivariate forecasting techniques to generate highly accurate demand predictions that consider both spatial correlations and temporal patterns. This comprehensive approach not only enhances forecast accuracy but also furnishes actionable insights for optimizing supply chain operations within the retail sector. Chapter 4, explores the significant challenge of estimating true consumer demand during out-of-stock events—a frequent and impactful issue in retail operations. Conventional demand forecasting methods often fail to capture the true essence of consumer desire, particularly when products are unavailable. This chapter introduces a sophisticated model that distinguishes between original demand, substitution demand, and deferral demand, thus providing a more accurate reflection of consumer intent and behavior. This detailed perspective on demand

Degree

thesis:*
Level thesis:degree_level
Doctoral
Department dc:contributor.department
Industrial and Systems Engineering
Grantor dc:publisher
Georgia Institute of Technology
Year dc:date.issued
2024

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Pan, Xinyue
Advisor dc:contributor.advisor
  • Montreuil, Benoit
Committee members dc:contributor.committeemember
  • Goldsman, David
  • Xie, Yao
  • Yang, Shihao
  • Agard, Bruno

Subjects

dc:subject × 6

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/1853/80046

Chain of custody

source
Harvested from
Georgia Tech
Base URL
repository.gatech.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Pan, Xinyue. Advanced Data-Driven Methodologies for Enhanced Demand Forecasting in Supply Chain Management. Doctoral thesis, Georgia Institute of Technology, 2024. https://hdl.handle.net/1853/80046