How to generate micro-agents? A deep generative modeling approach to population synthesis

Borysov, Stanislav S.; Rich, Jeppe; Pereira, Francisco C.

doi:10.1016/j.trc.2019.07.006

Cited by 70 publications

(41 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some SL methods are hindered by the scalability problem when the number of attributes is large. For Bayesian Networks and Hierarchical Mixtures, it is preferable for the number of attributes to be small, as pointed out by Borysov et al (51) and Zhang et al (50). In fact, a Bayesian network with six nodes (corresponding to attributes) contains some 3 million possible DAGs, and this number becomes 1.1 billion with a network of seven nodes (21).…”

Section: Comparison Of Methodsmentioning

confidence: 99%

“…Another SL-based method applied to population synthesis is the Bayesian Network (BN) ( 21, 50 ). A BN is a probabilistic graphical representation of the factorization of the joint probability distribution into conditional distributions, presented by means of a directed acyclic graph (DAG) ( 51 ). Sun and Erath ( 21 ) argued that a BN, by abstracting the structure of population systems using a DAG and local conditional probabilities, is able to capture and reproduce the complex dependence and higher-order interactions among a large set of variables.…”

Section: Statistical Learningmentioning

confidence: 99%

“…However, as the number of attributes increases, the HM method becomes very challenging to apply. Another challenge with this approach pertains to its robustness with respect to the selection of latent variables ( 51 ).…”

Section: Statistical Learningmentioning

confidence: 99%

“…More recently, Borysov et al ( 51 ) proposed a deep generative modeling (DGM) approach based on a variational autoencoder (VAE). According to these authors, each person is represented by a vector of random variables.…”

Section: Statistical Learningmentioning

confidence: 99%

“…SL methods can be applied first to construct a sample of suitable size, and afterward SR methods can be applied to create a population matching the marginals. Following this line, Borysov et al ( 51 ) in Denmark and Sun et al ( 13 ) in Singapore generated a population from household travel survey samples corresponding to 2.5% and 1% of the total population, respectively.…”

Section: Decision-making Proceduresmentioning

confidence: 99%

See 4 more Smart Citations

Comparing Methods for Generating a Two-Layered Synthetic Population

Yameogo

Gastineau

Hankach

et al. 2020

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

Synthetic population is used in many transport models ranging from trip-based, hybrid trip, tour-based, and activity-based models. As mobility decisions depend on both individuals’ characteristics and family situation, generating a two-layered population that takes into account not only the individual level but also household level is essential. In the literature, three main categories of methods for two-layered population generation have been proposed. These categories are synthetic reconstruction (SR), combinatorial optimization (CO), and statistical learning (SL). SR and CO methods produce synthetic populations by means of replicating individuals, whereas SL methods generate a population following a joint probability estimation. However, selecting a generation process is not straightforward as it depends on input data and synthetic population characteristics. To the best of our knowledge, no clear methodology for selecting between these methods exists. The main objectives of this paper are to provide (1) a detailed description of the available methods, (2) a comparison between these methods, and (3) a decision-making procedure for selecting between these methods. The description and comparison of the methods relies on different criteria: marginals availability, sample size, number of potential attributes that can be handled, population size to generate, possibility of zero-cell problem, and so forth. The advantages and shortcomings of each method are illustrated, and method performance is assessed. The decision-making procedure is carried out through the proposal of a decision tree. Researchers and practitioners have now access to a comprehensive and unified framework to select the appropriate method depending on available data and features of their modeling purposes.

show abstract

Section: Comparison Of Methodsmentioning

confidence: 99%