Prerequisites: CSE 458.
The subject follows up Bayesian Modelling by
addressing the machine learning of Bayesian Networks for
data mining. This has become a hot topic because of the
difficulty of eliciting Bayesian Networks from domain experts.
- Can we learn causal structure from correlational data?
- Bayesian confirmation theory
- Algorithms for learning causal structure
- How to estimate probabilities
- Linear and discrete Bayesian networks
- Integrating prior knowledge with data
- Stochastic sampling methods
- Metric learning with MML and alternatives
- Evaluating machine learning methods