A Novel Machine Learning Workflow to Classify Mobile Home Parks at Scale

Clinton Stipek; Sameer Nathawat; Taylor Hauser; Nagendra Singh

doi:10.1109/ACCESS.2025.3615118

A Novel Machine Learning Workflow to Classify Mobile Home Parks at Scale

Geospatial Data Modeling

Research output: Contribution to journal › Article › peer-review

Abstract

Understanding the built environment is essential to the overall study of population dynamics, grid infrastructure, emergency response, among others. In the United States there are multiple classifications for buildings within the built environment such as residential, signifying family homes while commercial buildings consist of apartments or larger structures which are multi-purpose. While there is a high level of understanding of where these aforementioned structures are located, there is a third class of structures, mobile home parks (MHP) which have been under-represented in the literature despite there being an estimated 2.7 million of them within the United States. Research has shown that individuals who reside in MHP are at higher risk to extreme events due to their location and structural integrity of residence. Attention must now turn to identifying MHP at scale to help first responders and policy makers understand where these at risk populations reside. To address for this gap, we develop a novel methodology to infer MHP at scale based off morphologies derived at a building level. We show that across 3 million buildings in 6 states within the United States it is possible to identify MHP with 83% accuracy. This novel approach to identify MHP from other structures within the built environment using a machine learning approach provides a new tool to leverage in relation to helping at-risk populations.

Original language	English
Pages (from-to)	172279-172292
Number of pages	14
Journal	IEEE Access
Volume	13
DOIs	https://doi.org/10.1109/ACCESS.2025.3615118
State	Published - 2025

Funding

This work was supported by the University of Tennessee (UT) – Battelle, Limited Liability Company (LLC), with U.S. Department of Energy (DOE) under Contract DE-AC05-00OR22725. This work was supported by the University of Tennessee (UT) – Battelle, Limited Liability Company (LLC), with U.S. Department of Energy (DOE) under Contract DE-AC05-00OR22725. The authors would like to thank Dr. Annetta Burger, Jack Gonzales, Joe Pyle, Darrell Roddy, Parker Irish, Brandyn Reynolds, and Zachary Barbose for their help and assistance. This work was supported by the University of Tennessee (UT) – Battelle, Limited Liability Company (LLC), with U.S. Department of Energy (DOE) under Contract DE-AC05-00OR22725. The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (https://www.energy.gov/doepublic-access-plan). The authors would like to thank Dr. Annetta Burger, Jack Gonzales, Joe Pyle, Darrell Roddy, Parker Irish, Brandyn Reynolds, and Zachary Barbose for their help and assistance. This work was supported by the University of Tennessee (UT) – Battelle, Limited Liability Company (LLC), with U.S. Department of Energy (DOE) under Contract DE-AC05-00OR22725. The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (https://www.energy.gov/doe-public-access-plan).

Keywords

Extreme events
knowledge discovery
machine learning
mobile home parks

Access to Document

10.1109/ACCESS.2025.3615118

Cite this

@article{b164c8763bb644cab39f7d7cf58688c2,

title = "A Novel Machine Learning Workflow to Classify Mobile Home Parks at Scale",

abstract = "Understanding the built environment is essential to the overall study of population dynamics, grid infrastructure, emergency response, among others. In the United States there are multiple classifications for buildings within the built environment such as residential, signifying family homes while commercial buildings consist of apartments or larger structures which are multi-purpose. While there is a high level of understanding of where these aforementioned structures are located, there is a third class of structures, mobile home parks (MHP) which have been under-represented in the literature despite there being an estimated 2.7 million of them within the United States. Research has shown that individuals who reside in MHP are at higher risk to extreme events due to their location and structural integrity of residence. Attention must now turn to identifying MHP at scale to help first responders and policy makers understand where these at risk populations reside. To address for this gap, we develop a novel methodology to infer MHP at scale based off morphologies derived at a building level. We show that across 3 million buildings in 6 states within the United States it is possible to identify MHP with 83\% accuracy. This novel approach to identify MHP from other structures within the built environment using a machine learning approach provides a new tool to leverage in relation to helping at-risk populations.",

keywords = "Extreme events, knowledge discovery, machine learning, mobile home parks",

author = "Clinton Stipek and Sameer Nathawat and Taylor Hauser and Nagendra Singh",

note = "Publisher Copyright: {\textcopyright} 2013 IEEE.",

year = "2025",

doi = "10.1109/ACCESS.2025.3615118",

language = "English",

volume = "13",

pages = "172279--172292",

journal = "IEEE Access",

issn = "2169-3536",

publisher = "Institute of Electrical and Electronics Engineers",

}

TY - JOUR

T1 - A Novel Machine Learning Workflow to Classify Mobile Home Parks at Scale

AU - Stipek, Clinton

AU - Nathawat, Sameer

AU - Hauser, Taylor

AU - Singh, Nagendra

PY - 2025

Y1 - 2025

N2 - Understanding the built environment is essential to the overall study of population dynamics, grid infrastructure, emergency response, among others. In the United States there are multiple classifications for buildings within the built environment such as residential, signifying family homes while commercial buildings consist of apartments or larger structures which are multi-purpose. While there is a high level of understanding of where these aforementioned structures are located, there is a third class of structures, mobile home parks (MHP) which have been under-represented in the literature despite there being an estimated 2.7 million of them within the United States. Research has shown that individuals who reside in MHP are at higher risk to extreme events due to their location and structural integrity of residence. Attention must now turn to identifying MHP at scale to help first responders and policy makers understand where these at risk populations reside. To address for this gap, we develop a novel methodology to infer MHP at scale based off morphologies derived at a building level. We show that across 3 million buildings in 6 states within the United States it is possible to identify MHP with 83% accuracy. This novel approach to identify MHP from other structures within the built environment using a machine learning approach provides a new tool to leverage in relation to helping at-risk populations.

AB - Understanding the built environment is essential to the overall study of population dynamics, grid infrastructure, emergency response, among others. In the United States there are multiple classifications for buildings within the built environment such as residential, signifying family homes while commercial buildings consist of apartments or larger structures which are multi-purpose. While there is a high level of understanding of where these aforementioned structures are located, there is a third class of structures, mobile home parks (MHP) which have been under-represented in the literature despite there being an estimated 2.7 million of them within the United States. Research has shown that individuals who reside in MHP are at higher risk to extreme events due to their location and structural integrity of residence. Attention must now turn to identifying MHP at scale to help first responders and policy makers understand where these at risk populations reside. To address for this gap, we develop a novel methodology to infer MHP at scale based off morphologies derived at a building level. We show that across 3 million buildings in 6 states within the United States it is possible to identify MHP with 83% accuracy. This novel approach to identify MHP from other structures within the built environment using a machine learning approach provides a new tool to leverage in relation to helping at-risk populations.

KW - Extreme events

KW - knowledge discovery

KW - machine learning

KW - mobile home parks

UR - https://www.scopus.com/pages/publications/105018116265

U2 - 10.1109/ACCESS.2025.3615118

DO - 10.1109/ACCESS.2025.3615118

M3 - Article

AN - SCOPUS:105018116265

SN - 2169-3536

VL - 13

SP - 172279

EP - 172292

JO - IEEE Access

JF - IEEE Access

ER -

A Novel Machine Learning Workflow to Classify Mobile Home Parks at Scale

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this