Enhancing operations management through descriptive, predictive, and prescriptive analytics

PDF
Zenouzzadeh_Dissertation.pdf - Published
Restricted to Staff of the repository only
Download (3MB)

URN:	urn:nbn:de:bsz:180-madoc-716164
Document Type:	Doctoral dissertation
Year of publication:	2025
Place of publication:	Mannheim
University:	University of Mannheim
Evaluator:	Prof. Dr. Raik Stolletz
Date of oral examination:	2025
Publication language:	English
Institution:	Business School > ABWL u. Produktion (Stolletz 2010-)
Subject:	650 Management
Keywords (English):	descriptive analytics , predictive analytics , prescriptive analytics , operations management
Abstract:	This dissertation examines how descriptive, predictive, and prescriptive analytics support decision-making and operational improvement in operations management and business analytics. Descriptive analytics is applied to a French call-center dataset under two business models to characterize customer patience, identifying advertisement-driven waves of call abandonment and an unexpected negative association between waiting time and abandonment consistent with a sunk-cost effect, with implications for richer queueing models. Predictive analytics is then advanced through an algorithm that estimates time-dependent queue dynamics by combining transient estimation, machine learning, and domain knowledge to improve interpretability and estimation accuracy. Finally, prescriptive analytics is developed for workforce scheduling in multi-department logistics systems via a mixed-integer programming formulation that optimizes intra-shift worker transfers to reduce costs and improve performance; a column-generation solution approach scales to large instances and demonstrates practical savings. Collectively, the dissertation shows how integrated analytics methods can generate insight, improve forecasting, and enable implementable optimization in service and logistics operations.