CSC2535 Spring 2008 - Lectures

LECTURE SCHEDULE: SUBJECT TO CHANGE

• January 9
  Lecture 1: Graphical Models 1 [notes]
  Required reading: Bishop, Chapter 8 pp 359-393
  

• January 16
  Lecture 2: Graphical Models 2 [notes]
  Required reading: Bishop, Chapter 8 pp 393-418
  

• January 23 (first assignment posted on web)
  Lecture 3: Variational Methods [notes]
  Required reading: Bishop, Chapter 10
  

• January 30
  Lecture 4: Sampling Methods [notes]
  Required reading: Bishop, Chapter 11
  

• February 6: First assignment due (at start of class)
  Lecture 5: Energy-based models
  notes as .htm , notes as .ppt , notes as .ps, 4 per page
  Required reading: Contrastive backpropagation [ps.gz] [pdf]
  Required reading: Energy-based models for sparse overcomplete representations. [ps.gz] [pdf]
  

• February 13 (second assignment posted on web)
  Lecture 6: New advances in deep belief nets
  notes as .htm , notes as .ppt , notes as .ps, 4 per page
  Required reading:
  

• February 20 No Lecture (reading week)
  

• February 27 Second assignment due (at start of class)
  Lecture 7: Models of words and documents [notes]
  Required reading: Latent Dirichlet allocation. [html]
  

• March 5
  Lecture 8: Missing data problems and applications [notes]
  Required reading: Learning from incomplete data. [pdf]
  Recommended reading: Collaborative filtering and the missing at random assumption. [pdf]

• March 12
  Lecture 9: Models for structured data (image and sequences) [notes]
  Required reading: Conditional random fields: Probabilistic models for segmenting and labeling sequence data. [pdf]
  Recommended reading: Learning and incorporating top-down cues in image segmentation. [pdf]

• March 19
  Lecture 10: Non-linear Dimensionality Reduction
  notes as .htm , notes as .ppt , notes as .ps, 4 per page
  Required Reading: ISOMAP article on web
  Required Reading: Local Linear Embedding [.pdf]
  Required Reading: Stochastic Neighbor Embedding [.ps]
  Optional Reading: Aspect Maps [.pdf]
  Required Reading: t-SNE [coming soon]
  

• March 26
  Lecture 11: Learning by maximizing agreement between outputs
  notes as .htm , notes as .ppt , notes as .ps, 4 per page
  Required Reading: A self-organizing neural network that discovers surfaces in random-dot stereograms. [pdf]
  Required Reading: Neighborhood Components Analysis [pdf]
  Required Reading: Learning a non-linear embedding by preserving class neighbourhood structure. [pdf]
  Optional Reading: Learning mixture models of spatial coherence [pdf]
  Optional Reading: Dimensionality Reduction by Learning an Invariant Mapping [pdf]
  

• April 2
  Lecture 12: Learning multiplicative interactions
  notes as .htm , notes as .ppt , notes as .ps, 4 per page
  Roland's lecture slides as .pdf
  Required reading: Tenenbaum and Freeman 2000 pdf]
  Required reading: Memisevic and Hinton 2007 pdf]
  

• April 9: Final Test (1.10pm-2.10pm)
  

• Tues April 15: Project due at Pratt 290D by 5.00pm
  

CSC2535 - Advanced Machine Learning: || www.cs.toronto.edu/~hinton/csc2535/