Quantitative Modelling of Social and Human Systems
Social systems and networks
We have studied different social systems in collaboration with social scientists: research in education, world knowledge about Chagas disease, simple model of coexistence of several currencies, or the users of the bike sharing system in Lyon.
Cliometrics and data science
We also conducts studies in Cliometrics, which is the mathematical study of economic historical time series.
In an interdisciplinary collaboration, we study different problems in the history of economics by developing new signal and physical methods for data processing
Geographical data analysis
We develop a multiscale wavelet-like framework for local multifractal analysis of geolocalized data, addressing challenges from sparse and heterogeneous supports. Applied to urban data, it reveals scale-free variations in major European cities, related to population heterogeneity.
Social data and complex systems
Inverse problem methods can be applied to social datasets (e.g., elections results, economic indexes...), to improve their understanding. For example, introducing similarity hypotheses for vote transfer matrices allows us to estimate their values, which cannot be observed, thus bringing information about the sociological determinants of the votes.
In geography, the geolocalized nature of data, together with their being defined on non dense heterogeneous supports, imply revisiting multiscale multivariate analysis prior to achieving contributions connected to state of the art material of interest to geographers.