- Presentation
- HAL publications
- Activity reports
GEOSTAT Research team
Geometry and Statistics in acquisition data
- Leader : Hussein Yahia
- Type : Project team
- Research center(s) : Bordeaux
- Field : Applied Mathematics, Computation and Simulation
- Theme : Optimization, Learning and Statistical Methods
Team presentation
GeoStat projecting makes fundamental and applied research on new non linear methods for the analysis of complex signals and systems, using paradigms and tools coming from the notions of scale invariance, predictibility, and the development of new formalisms such as the Multiscale Microcanonical Formalism.Research themes
GeoStat's research thematics are centered on the following theoretical developments :- Multiscale methods developped in Physics for the analysis of complex systems (computation of singularity exponents, Lyapunov exponents and large deviations, various definitions of entropy etc.),
- Predictibility in complex systems,
- Optimal wavelet decomposition,
- Analysis, classification, detection,
- Analysis of complex and turbulent signals in earth observation and remote sensing,
- Digital implementation of adaptative optics in astronomy,
- Speech analysis.
International and industrial relations
GeoStat is working in close collaboration with the following teams:- LATT (Laboratory for Astrophysics of Toulouse-Tarbes), UMR CNRS 5572, Toulouse, France .
- ICM-CSIC, Department of physical oceanography, Barcelona, Spain.
- LEGOS Laboratory, UMR CNRS 5566, Toulouse, France.
- Laboratory of theoretical physics and condensed matter University Paris 6, CNRS UMR 7600, Paris, France.
- IRIT, UMR CNRS 5505, Toulouse, France.
Keywords: Signal processing Non-linear methods Comple systems Turbulence Multiscale methods Complexity Scale invariance Complex signals Speech processing Adaptative optics Remote sensing Earth observ
Research teams of the same theme :
- CLASSIC - Computational Learning, Aggregation, Supervised Statistical, Inference, and Classification
- DOLPHIN - Parallel Cooperative Multi-criteria Optimization
- MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems
- MODAL - MOdel for Data Analysis and Learning
- REALOPT - Reformulations based algorithms for Combinatorial Optimization
- SELECT - Model selection in statistical learning
- SEQUEL - Sequential Learning
- SIERRA - Statistical Machine Learning and Parsimony
- TAO - Machine Learning and Optimisation
Contact
Team leader
Hussein Yahia
Tel.: +33 5 24 57 41 38
Secretariat
Tel.: +33 5 24 57 40 53
Inria
Inria.fr
Inria Channel

Find out more
See also