Supervisor

David McQuaid

Programme

MSc in Data Analytics

Subject

Computer Science

Abstract

Accurate demand forecasting is critical in operational settings where resource allocation and planning decisions depend on anticipated service volumes. Transactional systems that capture timestamped records provide valuable data sources for developing demand forecasts. This study examines hourly call volume forecasting using New Orleans police calls for service data, comparing the performance of statistical models, tree-based methods, and recurrent neural networks.

The research evaluates four primary modelling approaches: ARIMA models representing traditional statistical methods, XGBoost and Random Forest as a tree-based ensemble technique, and Gated Recurrent Units (GRUs) as deep learning alternatives. A naive seasonal model serves as the baseline benchmark. To ensure practical relevance, a robust evaluation framework tests the best-performing model from each category across six expanding time blocks, with each model trained and validated over successive two-week periods that mirror real operational deployment scenarios.

Results demonstrate that GRUs achieve the highest forecasting accuracy, with an R² of 0.74 and MAPE of 13.5%. XGBoost produces comparable performance, while ARIMA models show inferior results. To provide operational insights beyond predictive accuracy, a Random Forest model is developed as an interpretable explanatory framework, identifying key drivers that influence forecasting performance and offering actionable intelligence for practitioners implementing demand forecasting systems.

Date of Award

2025

Full Publication Date

2025

Access Rights

open access

Document Type

Capstone Project

Resource Type

thesis

Included in

Data Science Commons

Share

COinS