Evaluating Machine Learning Predictions for Budget Overruns in U.S. Mass Transit Projects - A Correlational Study - Dissertation Manuscript

Aug 2, 2026

This study evaluates supervised and unsupervised machine learning approaches to determine whether statistical correlations can validate U.S. mass transit project budget forecasts. Capital budget estimates for new projects have often been underestimated, creating budget overruns that complicate Federal Transit Administration funding decisions and force local transit authorities to obtain additional financing. Data from 108 projects drawn from publicly available U.S. government documents were modeled using SPSS with a neural network and k-means clustering, producing high variance explanations (R2=0.946 and R2=0.882). Results support an AI-driven, reproducible budget validation process that incorporates project-specific and environmental variables while improving transparency versus custom code.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages159

Reading Time6 h 41 min

AccessFree to download

Abstract

Chapter 1: Introduction

Statement of the Problem

Purpose of the Study

Introduction to the Conceptual Framework

Introduction to the Research Methodology and Design

Research Questions

Significance of the Study

Definition of Key Terms

Summary

Chapter 2: Literature Review

Literature Review Strategy

Conceptual Framework

Project Budget Methodology

U.S. Mass Transit Systems

Machine Learning

Data Management

Applying Machine Learning to Project Management

Additional Algorithm Types

Using SPSS for Machine Learning Research

Addressing Bias in Machine Learning

Summary

Chapter 3: Research Method

Research Methodology and Design

Population

Instrumentation

Study Procedures

Data Analysis

Machine Learning Using Supervised Learning

Machine Learning Using Unsupervised Learning

Assumptions

Evaluating Machine Learning Predictions for Budget Overruns in U.S. Mass Transit Projects: A Correlational Study Dissertation Manuscript Submitted to National University School of Business and Economics in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF BUSINESS ADMINISTRATION by PAUL ARTHUR BOUDREAU San Diego, California Abstract This study examined supervised and unsupervised machine learning methods to determine if statistical correlations could be established to validate mass transit project budgets. Capital budget forecasts for new mass transit projects in the United States have consistently been underestimated. These budget overruns resulted in funding allocation challenges for the Federal Transit Administration and obligated local transit authorities to secure additional financing. Developing accurate budget estimates is a significant challenge for project managers, particularly given the influence of cognitive biases such as optimism bias. Data from 108 projects were collected from publicly available U.S. government documents to serve as input for the machine learning models. Using a neural network and k-means clustering in SPSS software, the study created models that explained 94 .6%(R2 = 0 .946) and 88 .2%(R2 = 0 . 882) of the variance in project budget outcomes. The findings highlighted the importance for organizations to apply AIbased machine learning technology to validate budget forecasts. This research built on prior research by applying machine learning methods that had been explored in other project contextsand problem domains. By employing SPSS software with configurable settings, rather than relying on customized Python code, the study enhanced the transparency and accessibility of machine learning regression analysis. The results demonstrated that a budget validation process could be developed using reproducible, data-driven predictive models that incorporate both project-specific and environmental variables. Table of Contents Chapter 1: Introduction ................................................................................................................... 1 Statement of the Problem .......................................................................................................... 3 Purpose of the Study ................................................................................................................. 4 Introduction to the Conceptual Framework .............................................................................. 5 Introduction to the Research Methodology and Design ........................................................... 7 Research Questions ................................................................................................................... 9 Significance of the Study ........................................................................................................ 10 Definition of Key Terms ......................................................................................................... 11 Summary ................................................................................................................................. 12 Chapter 2: Literature Review ........................................................................................................ 13 Literature Review Strategy ..................................................................................................... 14 Conceptual Framework ........................................................................................................... 16 Project Budget Methodology .................................................................................................. 20 U.S. Mass Transit Systems ..................................................................................................... 33 Machine Learning .......................................................................

FAQs

What problem does the dissertation address in U.S. mass transit projects?

It addresses persistent underestimation of capital budget forecasts, which leads to budget overruns and funding allocation challenges for federal and local transit stakeholders.

Which machine learning techniques were used, and how were they implemented?

The study used a neural network and k-means clustering implemented in SPSS software, using data from 108 projects sourced from publicly available U.S. government documents.

How do the results contribute to improving budget estimation and validation?

The findings show that reproducible, data-driven predictive models can validate budget forecasts by incorporating both project-specific and environmental variables, supporting an AI-based validation process.

Aug 2, 2026

Upload by Valentina