Simulation and Bayesian networks

Unit Coordinator: Pere Puig, Rosa Camps, Rosario Delgado, Juan Ramon Gonzalez
Programme: Erasmus Mundus
ECTS Credits: 6
Semester: 1
Year: 2
Campus: Autonomous University of Barcelona
Language: English
Aims:

The objective of this course is to learn the methodologies of data simulation, bootstrapping, and permutation tests, that allow a fast solution to complex statistical problems without a deep knowledge of the classical statistics topics.

Bayesian networks as the graphical structures for representing probabilistic relationships among many variables, and their application to inference will be introduced.

Another objective is to learn the principles of causal inference.

Content:
  • Introduction to R.
  • Visualization of large-scale datasets with R.
  • Basics of Bayesian Networks.
  • Causal networks and Inference in Bayesian networks.
  • Learning Bayesian network parameters.
  • Data Simulation, Boostrapping and Permutation testing.
  • Permutation tests.
  • Jackknife.
  • Parametric Bootstrap.
  • Non-parametric Bootstrap.
Pre-requisites:

Elementary knowledge in Probability Theory and Statistical Inference.

Reading list:
  • Resampling methods: a practical guide to data Analysis. Phillip I. Good, 2006.
  • The jackknife, the bootstrap and other resampling plans. Bradley Efron, 1982.
  • Bootstrap methods and their application. A.C. Davison, D.V. Hinkley, 1997.
  • Learning Bayesian Networks, R. E. Neapolitan, Prentice Hall Series in Artificial Intelligence, 2004.
  • Probabilistic Methods for Bioinformatics with an Introduction to Bayesian Networks,  R. E. Neapolitan, Elsevier, 2009.

Print  

Related Articles