KTH Mathematics |

A Graduate Course: Bayesian Networks (7.5 p) FSF 3970 This course is of interest for engineers, statisticians and computer scientists who work with,e.g., modelling of highly complex systems, signal processing, data mining, artificial intelligence, robotics, or need understanding of statistical models using probabilities factorized according to directed acyclic graphs (DAGs) and the algorithms for the updating of probabilistic uncertainty in response to evidence, and statistical learning of model parameters and structures.
- Causality and directed acyclic graphs, and d-separation, conditional independence
- Markov properties for directed acyclic graphs and faithfulness.
- Learning about probabilities
- Structural learning; MDL, predictive inference
- Exponential familes and graphical models (Conditional Gaussian distributions)
- Causality and intervention calculus
- Chordal and decomposable graphs, moral graphs, junction trees, triangulation
- Local computation on the junction tree, marginalization operations propagation of probability and evidence, consistency
- Factor graphs, The Sum -Product algorithm (Wiberg's algorithm)
- T. Koski & J.Noble: Bayesian Networks and Causal Probability Calculus. 2009.Bayesian Networks: An Introduction (2009) published by Wiley. It may be ordered from amazon.co.uk here
- handouts
FIRST LECTURE: Friday, april 10th of 2015 at 14.15 - 16.00 in room: seminarierummet 3733 (room 3733 7th floor), institutionen för matematik, KTH, Lindstedtsvägen 25. You need to wait for entry outside the gated door LECTURES: Fridays at 14.15-16.00 Room: seminarierummet 3733 (room 3733 7th floor), institutionen för matematik, KTH, Lindstedtsvägen 25.
- TBA
- TBA
Course schedule and information Timo Koski Lecturer and Examiner Address: Department of Mathematics Royal Institute of Technology SE-100 44 Stockholm Sweden Email: tjtkoski@kth.se Phone: +46-8-790 71 34 Office: 3444 |

Published by: Timo Koski Updated: 30/8-2007 |