.png){kind=icon}
.png){kind=icon}
Return
Population synthesis: a problem-based review
.png){kind=icon}
PRE
AI
DOI：10.1080/01441647.2025.2469069.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Several studies have reviewed Population Synthesis (PopSyn) within Activity-Based Modelling (ABM) using a method-based approach. While this highlights progress in PopSyn development, it complicates the identification and comparison of specific challenges. This paper presents a comprehensive problem-based review of PopSyn, highlighting the critical challenges PopSyn faces. Four major issues are identified through a systematic review of the literature: data limitations (quality and quantity of input data), population heterogeneity (maintenance of population diversity), the curse of dimensionality (scalability), and adaptability (customisation and transferability). The review emphasises the need for greater focus on household relationship heterogeneity and model adaptability, which are crucial for accurate and practical PopSyn applications but are under-researched. It also underscores the importance of incorporating diverse data sources (part of data limitations) in the era of big data. By shifting from a method-based to a problem-based classification, this review aims to bridge the gap between academic research and practical application. This approach highlights existing gaps and challenges, provides a pathway for future research, and lays the groundwork for a comprehensive benchmark to assess various PopSyn methods. Ultimately, it aims to advance the field and promote broader adoption in real-world scenarios.
Keywords:
Activity-based
population synthesis
demand modelling
problem-based
reclassification
artificial population