Keybioeffects

Biostatistics and Ecological Modelling

Biostatistics and Ecological Modelling

Title                    

Biostatistics and Ecological Modelling (International Study Programme: CNRS, Toulouse)

This title was changed during the last university reform and was transformed to the three following courses:

 
Title Methods for Estimation, Optimization and Learning (International Study Programme: CNRS, Toulouse)
Institute responsible for organization CNRS (Sovan Lek)
Objectives Provide to the students all statistical methods for estimations and optimizations of parameters, and learning methods to pattern, classification and prediction of complex ecological data.
Contents

- Estimations : Bayesian, AIC, Maximum likelihood
- Markov decisional processes
- Simulated annealing
- Fuzzy logic
- Machine learning techniques:
     ·SVM, CART algorithms
     ·ANN algorithms
- Genetic algorithms

Background necessary Elementary statistic and probability
Place University of Toulouse (France)
Duration (h) 60h
ECTS 6
Time September to Dicember
Max number of participants 10
Title Geostatitics and time series (International Study Programme: CNRS, Toulouse)
Institute responsible for organization CNRS (Sovan Lek)
Objectives Provide students statistical techniques to model time series data and spatial data.
Contents

Temporal series

  - Probability remain
  - Basic concept for signal processing
  - Time series data
  - Kalman filter, estimation of random signal
  - Linear signal processing
  - Particular algorithm, Monté-Carlo method
  - Filtering by genetic algorithms

Spatial analyses

  - Basic concept of spatial data analyses
  - Spatial statistics: covariance, spatial autocorrelation…
  - Neighboring distance
  - Ripley function, dispersion indices, spectral analyses

Background necessary Basic statistic, probability and mathematic
Place University of Tolouse (France)
Duration (h) 60h
ECTS 6
Time September to Dicember
Max number of participants 10
Title Modelling and numerical analyses (International Study Programme: CNRS, Toulouse)
Institute responsible for organization CNRS (Sovan Lek)
Objectives Provide to students techniques for numerical analyses to simulate ecosystem functioning, using several dynamic models.
Contents - Deterministic models continuous and discrete (Ordinary Differential Equation ODE).
- Stochastic effects and their impacts on the functioning of the dynamic system.
- Temporal statistical models (Discrete and continuous Markov Chain)
- Modelling spatio-temporal, deterministic and stochastic (partial differential equations PDE, individual Based Model IBM).
- Numerical resolution of the models.
Background necessary Basic statistic, probability and mathematic
Place University of Tolouse (France)
Duration (h) 60h
ECTS 6
Time September to Dicember
Max number of participants 10