Skip to content

Bayesian Models and Machine Learning

2 ECTS Kourou Year 2 Semester 3 Elective Tool


The Bayesian framework allows to build flexible models, to explicitly model uncertainty, and to account for prior knowledge about the processes and parameters. The basics of Bayesian data analysis will be presented, with application with R and Stan on different dataset related to tropical ecology.

The machine learning part focuses on recent advances in machine learning tools that can be applied to ecological models. Different classification methods will be presented and the basics of deep learning and convolution neural networks methodologies. Again, R will be used to apply the methods to ecological data.

Learning outcomes

After completing the course, the student should be able to analyse a dataset, to build, write and implement a model with Rstan, to interpret the results, and to be aware of the limitations of these methods. They should also be able to apply machine learning algorithms adapted to different tasks (classification, segmentation…) and to wisely split the data set into test, training and validation sets.

Teaching and learning methods

The course is entirely provided in the classroom.

Type of assessment

  • Individual
  • Written assignment
  • A small dataset will be provided to the students and they will have to conduct an analysis on it using the Bayesian framework.


  • Location : Kourou
  • ECTS granting : AgroParisTech
  • Organisation : AgroParisTech


Mélaine Aubry-Kientz

Mélaine Aubry-Kientz

  • AgroParisTech, Kourou France