Decoding Urban-health Nexus - Interpretable Machine Learning Illuminates Cancer Prevalence based on Intertwined City Features

Aug 4, 2026

Upload by Himbo

This study investigates the interplay among social demographics, built environment characteristics, and environmental hazard exposure features in determining community-level cancer prevalence. Using data from five U.S. Metropolitan Statistical Areas (Chicago, Dallas, Houston, Los Angeles, and New York), an XGBoost interpretable machine learning model predicts cancer prevalence and ranks influential features. Results highlight age, minority status, and population density, while causal-evidence experiments indicate that expanding green space and reducing developed areas and total emissions may alleviate prevalence.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages24

Reading Time1 h 0 min

AccessFree to download

Abstract

Introduction

Public health outcomes and urban feature interactions

Cancer burden and need to uncover contributing factors

Built environment effects and environmental hazards

ar iv: 30 . 1847 [ cs . LG] 22 Jun 2023 Decoding Urban-health Nexus: Interpretable Machine Learning Illuminates Cancer Prevalence based on Intertwined City Features Chenyue Liu 1*, Ali Mostafavi2 1Ph.D. Student, Urban Resilience.AI Lab, Zachry Department of Civil and Environmental Engineering, Texas A&M University, College Station, United States; e-mail: [liuchenyue@tamu.edu](liuchenyue@tamu.edu) 2Associate Professor, Urban Resilience.AI Lab Zachry Department of Civil and Environmental Engineering, Texas A&M University, College Station, United States; e-mail: [amostafavi@civil.tamu.edu](amostafavi@civil.tamu.edu) Abstract This study investigates the interplay among social demographics, built environment characteristics, and environmental hazard exposure features in determining communitylevel cancer prevalence. Utilizing data from five Metropolitan Statistical Areas in the United States — Chicago, Dallas, Houston, Los Angeles, and New York—the study implemented an XGBoost machine learning model to predict the extent of cancer prevalence and evaluate the importance of different features. Our model demonstrates reliable performance, with results indicating that age, minority status, and population density are among the most influential factors in cancer prevalence. We further explore urban development and design strategies that could mitigate cancer prevalence, focusing on green space, developed areas, and total emissions. Through a series of experimental evaluations based on causal inference , the results show that increasing green space and reducing developed areas and total emissions could alleviate cancer prevalence. The study and findings contribute to a better understanding of the interplay among urban features and community health and also show the value of interpretable machine learning models for integrated urban design to promote public health. The findings also provide actionable insights for urban planning and design, emphasizing the need for a multifaceted approach to addressing urban health disparities through integrated urban design strategies. Key words: Urban Health, Interpretable Machine Learning, Causal inferences, Integrated Urban Design, Sustainability, Environmental Justice 1. Introduction Public health outcomes in cities arise from the interplay of complex and nonlinear interactions among urban features. Yet, the existing approaches for formulating public health policies in cities focus primarily on a limited number of features using statistical methods that assume linear relationships between features and health outcomes. This limitation has hindered integrated urban design strategies to reduce disease prevalence in cities by failing to consider the salutary effects of improving the built environment and environmental hazard characteristics in different areas of cities. To address this gap, in this study, we examine the extent to which urban features related to the built environment, socio demographic characteristics, and environmental hazards and their non-linear interactions shape the prevalence of cancer using interpretable machine learning models. The incidence of cancer and its associated morbidity and mortality rates pose significant public health challenges worldwide. The American Cancer Society estimates that there will be more than 1.9 million new cancer cases diagnosed and more than 600,000 cancer deaths in the United States in 2023 alone [1] . While these statistics are indeed alarming, they also underscore the urgent need for uncovering the factors contributing to cancer prevalence to inform preventative strategies. Despite the substantial advancements in understanding the genetic [2] [3] [4] and lifestyle factors [5] [6] associated with cancer, less attention has been given to the potential role of the built environment and environmental hazard exposures. Recent research, however, is beginning to highlight the significant effects that social demographics [7], urban de

FAQs

What factors does the study examine in relation to cancer prevalence?

It examines social demographics, built environment characteristics, and environmental hazard exposure features and how their non-linear interactions shape community-level cancer prevalence.

Which machine learning approach is used, and what does it provide?

An XGBoost model is used to predict the extent of cancer prevalence and evaluate the importance of different features, supporting interpretability of the results.

What interventions does the study suggest to mitigate cancer prevalence?

Causal inference-based evaluations indicate that increasing green space and reducing developed areas and total emissions could alleviate cancer prevalence.

Aug 4, 2026

Upload by Himbo