Generating a synthetic population of individuals in households: Sample-free vs sample-based methods

TL;DR

This paper compares sample-free and sample-based methods for generating synthetic household populations, finding that the sample-free approach better fits reference data and is less data-dependent, despite requiring more pre-processing.

Contribution

It provides a comparative analysis of two methods for synthetic population generation, highlighting the advantages of the sample-free approach in accuracy and data requirements.

Findings

Sample-free method better fits reference distributions.

Sample-free method is less data demanding.

Sample-based method highly depends on initial sample quality.

Abstract

We compare a sample-free method proposed by Gargiulo et al. (2010) and a sample-based method proposed by Ye et al. (2009) for generating a synthetic population, organised in households, from various statistics. We generate a reference population for a French region including 1310 municipalities and measure how both methods approximate it from a set of statistics dervied from this reference population. We also perform sensitivity analysis. The sample-free method better fits the reference distributions of both individuals and households. It is also less data demanding but it requires more pre-processing. The quality of the results for the sample-based method is highly dependent on the quality of the initial sample.