| Date |
Topic |
Read |
Assignment
Due |
| Jan |
|
| 15 |
Course Overview |
|
|
| 20 |
no class |
|
|
| 22 |
Introduction: Key Concepts and Problems |
Ch. 1 |
|
| 27 |
Probability Review: Probability Theory |
|
|
| 29 |
Probability Review: Probability Distributions |
Ch. 2 |
|
| Feb |
|
| 3 |
Linear Models for Regression |
Ch. 3 |
Problem set 1 |
| 5 |
Bayesian Linear Regression |
|
|
| 10 |
Linear Models for Classification: Discriminant
Functions |
Ch. 4 |
|
| 12 |
Linear Models (cont): Logistic
Regression, Laplace Approx. |
|
|
| 17 |
substitute
Monday schedule |
|
|
| 19 |
Classification methods (cont.) |
|
Program 1: Linear methods |
| 24 |
Mid-term: in class |
|
|
| 26 |
Kernel Methods: Dual representations,
radial basis functions |
Ch. 6 |
|
| Mar |
|
| 3 |
Kernel Methods: Gaussian Processes |
|
|
| 5 |
Sparse Kernel Machines: SVM, RVM |
Ch. 7 |
|
| 7-15 |
spring
recess |
|
|
| 17 |
Graphical Models: Bayes Nets |
Ch. 8 |
Problem set 2
|
| 19 |
Markov Random Fields
|
|
|
| 24 |
Inference algorithms for graphical models
|
|
|
| 26 |
K-means clustering, mixtures of Gaussians,
EM |
Ch. 9 |
|
| 31 |
Basic sampling algorithms
|
Ch. 11 |
Program 2: Kernel Methds |
| Apr |
|
| 2 |
Markov Chain Monte Carlo, Gibbs sampling
|
|
|
| 7 |
Sequential data: hidden Markov models
|
Ch. 13
|
|
| 9 |
Sequential data: particle filtering
|
|
|
| 14 |
Combining methods: Boosting, tree-based
methods
|
Ch. 14
|
Program 3: Graphical models |
| 16 |
Combining methods: conditional mixtures
|
|
|
| 21 |
Continouous latent variables: principal
components
|
Ch. 12 |
|
| 23 |
substitute
Monday schedule
|
|
|
| 28 |
ICA, Kernel PCA
|
|
Program 4: Combining Methods |
| 30 |
Take home exam due
|
|
|