Stochastic modelling is used in fundamental research to summarise the properties of complex systems but also in decision making as a forecasting tool or to build evolution scenarios. Statistical methods based on independent samples or projective methods of data analysis are often too simple to address engineering and management problems. The aim of this course is an introduction to models with a stochastic component, that may represent a really stochastic phenomenon (Brownian motion in pollutant diffusion), or a situation where the factors are too numerous or the measures not feasible (seeds dispersion by the wind, by animals).
First week: Introduction. Point processes and application to the position of trees in a population. Markov chains.Second week: Application of Markov chains.Time series. Geostatistics.Third week: Projects: Markov chains and genetic codes. Stochastic geometry and hydrographical net. Forecasting in time series.
Evaluation mechanism : Project reports
Last Modification : Friday 27 February 2009