View in EDS

Sydney’s residential relocation landscape: Machine learning and feature selection methods unpack the whys and whens

Saved in:
Bibliographic Details
Title: Sydney’s residential relocation landscape: Machine learning and feature selection methods unpack the whys and whens
Authors: Bostanara, Maryam, Siripanich, Amarin, Ghasri, Milad, Rashidi, Taha
Source: Journal of Transport and Land Use, Vol 17, Iss 1 (2024)
Publisher Information: Center for Transportation Studies, 2024.
Publication Year: 2024
Subject Terms: anzsrc-for: 33 Built Environment and Design, 4406 Human Geography, 3304 Urban and Regional Planning, anzsrc-for: 40 Engineering, anzsrc-for: 1604 Human Geography, Machine learning, 0502 economics and business, 11. Sustainability, Machine Learning and Artificial Intelligence, Logistics and Supply Chains, anzsrc-for: 44 Human Society, anzsrc-for: 4406 Human Geography, 4005 Civil Engineering, Tourism and Services, 33 Built Environment and Design, 40 Engineering, 44 Human Society, HE1-9990, anzsrc-for: 1507 Transportation and Freight Services, TA1001-1280, anzsrc-for: 35 Commerce, 05 social sciences, anzsrc-for: 1205 Urban and Regional Planning, 3509 Transportation, 35 Commerce, Survival analysis, 15. Life on land, anzsrc-for: 3304 Urban and Regional Planning, Residential self-selection, Accessibility, anzsrc-for: 3509 Transportation, Management, Residential relocation, Transportation engineering, Networking and Information Technology R&D (NITRD), anzsrc-for: 4005 Civil Engineering, Transportation and communications
Description: This study investigates household residential relocation timing, an aspect vital for transport and urban planning. Analyzing a high-dimensional dataset from 1,024 relocations in Sydney, Australia, the research contrasts ten machine learning survival techniques with three classical survival models. Results indicate that when classical models are paired with tree-based automated feature selectors, they align closely with machine learning outcomes. Notably, the GBM, XGBoost, and Random Forest models emerge as standout performers. The study provides a comprehensive comparison between automatic and manual feature selection, shedding light on variables influencing households’ duration of stay. While stacked ensemble modeling, which leverages predictions from various models, is used to enhance accuracy, the improvements are marginal, underscoring inherent modeling challenges, particularly the recurring issue of misclassifying specific pairs of households in the concordance index measure. A thorough feature analysis highlights homeownership as the foremost predictor, underscoring the importance of recent life events and accessibility features in relocation decisions. The research emphasizes the importance of considering the accessibility of both current and future homes in relocation models, with 20% feature significance in model outcomes. Building on these foundational insights, the study paves the way for a deeper understanding of individual decision-making processes in sustainable urban planning.
Document Type: Article
File Description: application/pdf
ISSN: 1938-7849
DOI: 10.5198/jtlu.2024.2440
DOI: 10.2139/ssrn.4221706
Access URL: https://doaj.org/article/60ff59766650416490912c137f4644c4
Rights: CC BY NC ND
CC BY
Accession Number: edsair.doi.dedup.....c7d7871ca1251d7410d32d2638419508
Database: OpenAIRE
Description
Abstract:This study investigates household residential relocation timing, an aspect vital for transport and urban planning. Analyzing a high-dimensional dataset from 1,024 relocations in Sydney, Australia, the research contrasts ten machine learning survival techniques with three classical survival models. Results indicate that when classical models are paired with tree-based automated feature selectors, they align closely with machine learning outcomes. Notably, the GBM, XGBoost, and Random Forest models emerge as standout performers. The study provides a comprehensive comparison between automatic and manual feature selection, shedding light on variables influencing households’ duration of stay. While stacked ensemble modeling, which leverages predictions from various models, is used to enhance accuracy, the improvements are marginal, underscoring inherent modeling challenges, particularly the recurring issue of misclassifying specific pairs of households in the concordance index measure. A thorough feature analysis highlights homeownership as the foremost predictor, underscoring the importance of recent life events and accessibility features in relocation decisions. The research emphasizes the importance of considering the accessibility of both current and future homes in relocation models, with 20% feature significance in model outcomes. Building on these foundational insights, the study paves the way for a deeper understanding of individual decision-making processes in sustainable urban planning.
ISSN:19387849
DOI:10.5198/jtlu.2024.2440