Imperial College London (CO-493)
The course is co-taught by Duncan Gillies and Marc Deisenroth.
Syllabus
- Bayes’ Theorem and Bayesian Inference
- Bayesian Decision Trees -Evidence and message passing
- Inference in singly connected networks
- Building networks from data
- Cause and Independence
- Model Accuracy
- Approximate Inference
- Exact Inference
- Probability propagation in Join Trees
- Probability distributions and parameter estimation
- Graphical models
- Gaussian mixture models, EM, model selection
- Dimensionality reduction with PCA
- Linear discriminant analysis
- Sampling
- Variational inference by Hugh Salimbeni