Offre de Post-Doc à pour voir dès septembre 2019.
Multifractal Analysis of urban/population/ressources data
Physics Department of ENS de Lyon
Available September 2019
Research themes : Multifractal analysis and modeling for cities.
Duration : 12 months, with possible 12-month renewal.
Advisors. Stephane Roux, firstname.lastname@example.org, perso.ens-lyon.fr/stephane.roux
Patrice Abry, email@example.com, perso.ens-lyon.fr/patrice.abry
Olivier Bonin, firstname.lastname@example.org, www.lvmt.fr/equipe/olivier-bonin/
Stephane Jaffard, email@example.com, perso-math.univ-mlv.fr/users/jaffard.stephane/
perso.ens-lyon.fr/patrice.abry/PostDocMultifracUrbanData.pdfPost-Doc position :
Required qualications : Candidates should have received a high-level academic forma-
tion in statistical signal processing. They must be strongly motivated by exploratory and
multidisciplinary research topics. Skills and practice of Matlab and Python are expected.
Location : The candidate will be hired by the University of Paris-Est through the project
AAP Impulsion MULTIFRAC, and will work with the Signal, System and Physics team,
within the Physics Laboratory at ENS Lyon.
Application : Interested candidates should send their CV, a motivation letter and at least
Keywords : Geography, population, urbanism, city growth, image processing, scale-free,
multifractal, point process, real-world data.
Description : Often, data associated with city organization or growth, such as population
densities, building structures, ressources are multiscale in nature, that is their dynamics
are spread across a wide range of scales, with no typical scale playing a specic role. The
analysis and modeling of such scale-free organizations and dynamics in city-based data
constitute a long standing research themes, with important stakes in term of territory
management and population behaviors. However, because data can be very dierent in
nature (density map, binary image of build area,...), in modality (teledetection data, si-
mulated data, road network, population density...) and resolution, their analyses requires
skills beyond the original-mother eld "geography" and calls for interdisciplinary teams
gathering expertize in computer sciences, mathematics and statistics.
Since the early 1960s, fractal concepts were proposed as models to account for scale-
free dynamics of multiscale phenomena and multifractal analysis has nowadays become a
standard tool in signal/image processing. However, the use of fractal concepts and tools to
analyze city-type data implies to address several critical issues stemming from the nature
and type of data, including heterogeneities, anisotropy, lacunarity,...
Therefore, the main goal of the current project will be to develop, assess and put at
work on real data tools permitting an eective, accurate and relevant analysis of scale-free
and multifractal properties in city-type data. These tools will need to take into account
the binary (point-process like) nature of some such data and also will be orientated toward
multivariate analysis (several images dierent in nature of the same area need to ne analy-
zed jointly). The project thus implies methodological contributions, numerical simulation
assessment and application to real data. The project is thus highly interdisciplinary and
will be conducted in collaboration with expert in urban/city organizations and structures
as well as with mathematicians and image processing teams.
(c) GdR 720 ISIS - CNRS - 2011-2019.