Synthetic populations for epidemic, demographic and mobility modeling

Davide Torre; Miguel Artigues; Benjamin Faucher; Roberto Alcaraz; Jacopo D'Ignazi; Antonia Tugores; Mattia Mazzoli; Daniela Paolotti; José J. Ramasco; Eugenio Valdano

arXiv · 2610.21784

Synthetic populations for epidemic, demographic and mobility modeling

Davide Torre·Miguel Artigues·Benjamin Faucher·Roberto Alcaraz·Jacopo D'Ignazi·Antonia Tugores·Mattia Mazzoli·Daniela Paolotti·José J. Ramasco·Eugenio Valdano

View original source ↗Read source document ↗Export metadata ↓

Abstract

Synthetic populations are essential for individual-based simulations, such as agent-based models, enabling the evaluation of policy interventions for epidemic control, demographic studies and mobility management, to mention a few. Existing synthetic populations are often designed for specific settings and rarely combine nationwide coverage with realistic and explicit spatial detail. This limits their ability to represent realistic spatial relationships and interplay between households, workplaces, leisure activities, and mobility patterns, reducing models utility for targeted decision-making. Here, we present a harmonized dataset of synthetic populations taking as case studies Spain, France, and Italy. Despite heterogeneous census data sources and formats, we release a unified data structure ensuring consistency and interoperability. The synthetic population generation pipeline reproduces household composition, population density, age-sex structure, and occupation distribution at census-tract level, enriched with municipality-level commuting flows and economic activity patterns. Country-specific pre-processing, statistical calibration, and data integration produce coherent demographic and spatial characteristics suitable for large-scale, country-agnostic simulations. As the pipeline relies on publicly available census data, it is readily extensible to other countries, supporting high-resolution modeling of infectious diseases, hazard exposure, demographic analysis, and urban mobility.

Explore related subjects

physics.soc-ph

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Davide Torre, Miguel Artigues, Benjamin Faucher, Roberto Alcaraz, Jacopo D'Ignazi, Antonia Tugores, Mattia Mazzoli, Daniela Paolotti, José J. Ramasco, Eugenio Valdano. 2026-10-15. Synthetic populations for epidemic, demographic and mobility modeling. https://arxiv.org/abs/2610.21784

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Pairwise assessment correlations predict cooperation of the Leading Eight under private assessment

Indirect reciprocity can sustain cooperation when individuals use reputations to decide whom to help. The Leading Eight norms sustain such cooperation under public assessment, where all individuals share the same assessments. Under private assessment, however, observers may disagree, yet their assessments remain correlated. Two quantities, the average fraction of good assessments $h$ and the conditional probability $h_{\mathrm G}$ that one observer assesses a player as good given that another does, are known to determine the stability of cooperation, but how $h_{\mathrm G}$ itself is determined has remained open. Within a pairwise approximation, we derive a closed equation for $h_{\mathrm G}$ that, together with that for $h$, determines both quantities for all eight norms, including third-order ones. This description substantially improves agreement with simulations over a mean-field approximation that neglects correlations, and recovers the known classification of the Leading Eight into three classes. The remaining discrepancies are largest for L2 and L5 at low error rates and decrease as assessment error increases. Relating $h_{\mathrm G}$ to the variance of individual goodness suggests that these discrepancies reflect information beyond the first two moments of the goodness distribution. Our framework provides a concise and analytically tractable tool for evaluating social norms under private assessment.

physics.soc-ph↗

arXiv2026

Higher-Order Co-Infection on Hypergraphs

People encounter infections through overlapping groups, such as households, classrooms, and workplaces. Understanding how these shared contacts affect the spread of multiple pathogens is critical to determining what information epidemic models need to retain. We introduce HyCo-SIS (Higher-order Co-infection SIS), a discrete-time SIS model on hypergraphs with explicit reinforcement between overlapping active groups and pairwise transmission. We investigate how this reinforcement affects infection dynamics and whether simplified representations of hypergraphs preserve the resulting behavior. We derive an exact representation of overlap reinforcement as target-dependent four-person interactions and determine when a symmetric representation is possible. Numerical experiments examine reinforcement strength, contact structure, and susceptibility to secondary infection, as well as calibrated pairwise approximations. Our results demonstrate that redistributing overlap weights can substantially change finite-time co-infection even when every individual's total overlap coefficient is preserved. We further derive an exact five-person calculation and demonstrate why equal mean infection pressure does not necessarily produce equal acquisition probabilities. These findings identify how the allocation of higher-order interactions can influence contagion dynamics beyond their aggregate strength. Beyond epidemiology, this work contributes to the statistical physics of spreading processes by examining how the arrangement of many-body interactions shapes collective dynamics, and what information is lost when those interactions are simplified.

physics.soc-ph↗

arXiv2026

Constrained modularity maximization in complex networks via branch-and-price

Modularity maximization is widely used to identify community structure in complex networks. In many applications, however, community detection must also incorporate prior knowledge and domain-specific requirements beyond network topology. We study constrained modularity maximization with restrictions on the number of communities, must-link and cannot-link relations, and general resource constraints. While fast heuristics can provide high-quality solutions, they generally lack optimality guarantees, whereas exact mathematical-programming approaches can become computationally demanding as the network size and constraint complexity increase. We develop a unified optimization framework offering complementary heuristic and exact solution methods. These include a constrained Louvain heuristic for feasible solution generation, a column-generation heuristic with graph-partitioning-based pricing, and exact branch-and-price algorithms using mixed-integer linear programming and sparse mixed-integer quadratic programming formulations for the pricing subproblem. We develop a cutting-plane procedure providing independent upper bounds to assess solution quality when exact optimization cannot be completed. Experiments on 11 unconstrained networks with up to 1,005 nodes and 30 constrained instances with up to 379 nodes demonstrate the complementary roles of these methods. The column-generation heuristic rapidly improves the constrained Louvain solutions across the tested constraint classes, while branch-and-price using the sparse quadratic formulation solves 29 of the 30 constrained instances to optimality within a one-hour time limit. This formulation becomes particularly effective on larger, sparse, undirected networks by substantially reducing the size of the exact pricing model. Overall, the framework guides method selection to balance computational time, solution quality, and optimality guarantees.

physics.soc-ph↗