Bayesian inference is a powerful and increasingly popular statistical approach. It can deal with complicated problems where classical frequentist analysis would be difficult to apply. Indeed, Bayesian data analysis and frequentist methods provide different ways to draw conclusions from data and address random variation. While classical frequentist approaches calculate the probability of observing data given a specific value of a parameter, Bayesian statistics provide conditional probabilities for the different values of a parameter given the data. In other words, Bayesian approach allows scientists to combine new data with their existing knowledge or expertise.
Firstly facing some computational challenges for large problems solving, recent advances in computational techniques and especially the discovery of Markov Chain Monte Carlo (MCMC) simulation methods have led to an explosion of interest in Bayesian statistics and modelling in many areas including computational biology and ecology.
The course will be centered on "bayesian data analysis" applied to biological problems. Topics addressed during this course include single-and multi-parameter bayesian models, hierarchical models and bayesian computation technics (MCMC).
This course is intended for life scientists who already have some good knowledge of statistics and the programming language "R".
At the end of this course, participants are expected to be able to:
- understand the difference between a frequentist and a bayesian approach
- compute the posterior density in a simple case
- understand the basics of hierarchical modeling
- get a basic idea of the different bayesian computational methods
- perform a bayesian analysis on real data