Machine Learning
Graphical Models
Bayesian Networks, examples, conditional
independence, inference
Marc Toussaint
FU Berlin
The need for modelling
• Given a real world problem, translating it to a well-defined learning
problem is non-trivial.
• The “framework” of plain regression/classification is rather restricted:
input x, output y.
• Graphical models (probabilstic models with multiple random variables
and dependencies) are a more general framework for modelling
“problems”; regression & classification become a special case;
Reinforcement Learning, decision making, but also language
processing, image segmentation, are special cases.
2/21
Graphical Models
• The core difficulty in modelling is specifying
What are the relevant variables?
How do they depend on each other?
(Or how could they depend on each other→ learning)
• Graphical models are a simple, graphical notation for
1) which random variables exist
2) which random variables are “directly coupled”
Thereby
de/variables/models/modelling/ran/dom/Graphical/pend/special/work/
de/variables/models/modelling/ran/dom/Graphical/pend/special/work/
-->