CS 6362 - Advanced Machine Learning




Schedule

Week 1 (August 23): Course Introduction, Bayesian Linear Regression
Wednesday: Course Introduction, Bayesian Linear Regression
Assignment 1 Posted
Required Reading: (Rasmussen & Williams) Ch. 1, 2.1
Week 2 (August 30): Gaussian Processes
Monday: Gaussian Process Regression
Required Reading: (Rasmussen & Williams) Ch. 2.2-2.6
Recommended Reading: Exploring Gaussian Processes
Wednesday: Gaussian Process Classification, pt. 1
Required Reading: (Rasmussen & Williams) Ch. 3.1-3.4
Friday:
Assignment 1 Due
Week 3 (September 6): Classification, Model Selection
Monday: Gaussian Process Classification, pt. 2, Model Selection Occam’s Razor notebook
Assignment 2 Posted
Required Reading: (Rasmussen & Williams) Ch. 3.5, 5.1-5.3
Wednesday: Model Selection keynote, pdf
Required Reading: (Rasmussen & Williams) Ch. 5.3-5.5
Week 4 (September 13): Approximation Methods
Monday: Nystrom method, Sparse GPs keynote, pdf
Assignment 2 Due
Required Reading: (Rasmussen & Williams) Ch. 8.1-8.4, Sparse GPs using Pseudo-inputs
Wednesday: Variational Gaussian Processes keynote, pdf
Assignment 3 Posted
Required Reading: Variational Learning of GPs, GPs for Big Data, Variational GPs for Classification
Week 5 (September 20): Covariance functions, Learning Theory
Monday: Covariance function properties, construction, and analysis keynote, pdf
Required Reading: (Rasmussen & Williams) Ch. 4
Recommended Reading: Kernel Cookbook
Wednesday: Learning Theory keynote, pdf
Required Reading: (Rasmussen & Williams) Ch. 7.1-7.5
Recommended Reading: PAC-Bayesian Margin Bounds
Week 6 (September 27): Latent Variable Models, Neural Networks
Monday: Gaussian Process Latent Variable Models keynote, pdf
Required Reading: GPLVM
Recommended Reading: Variational Inference for GPLVM
Wednesday: GPs and Neural Networks keynote, pdf
Required Reading: Deep Neural Networks as Gaussian Processes, GPs and Learning Dynamics
Friday:
Assignment 3 Due
Week 7 (October 4): Other Models, Midterm
Monday: Connections to other models, Midterm Review keynote, pdf
Required Reading: (Rasmussen & Williams) Ch. 6
Wednesday: Midterm
Week 8 (October 11): Project Proposals, Multi-Task Learning
Monday: Project Proposal Presentations
Wednesday: Multi-Task Learning keynote, pdf
Required Reading: Multilinear Multitask Learning, Learning Multiple Tasks using Manifold Regularization
Week 9 (October 18): Multi-Task Learning, Domain Adaptation
Monday: Multi-Task Gaussian Processes keynote, pdf
Required Reading: Multi-task Gaussian Process Prediction
Wednesday: Domain Adaptation: Overview keynote, pdf
Required Reading: Frustratingly Easy Domain Adaptation, Metric Learning for Visual Domain Adaptation
Week 10 (October 25): Domain Adaptation
Monday: Subspace-based Domain Adaptation keynote, pdf
Required Reading: Subspace Alignment for Domain Adaptation, Domain Invariant Projection
Wednesday: Deep Domain Adaptation keynote, pdf
Required Reading: Unsupervised Domain Adaptation by Backpropagation, CyCADA
Week 11 (November 1): Online Learning, Active Learning
Monday: Online Learning keynote, pdf
Assignment 4 Posted
Required Reading: Online learning and stochastic approximations
Recommended Reading: Stochastic Variational Inference
Wednesday: Active Learning: Overview keynote, pdf
Required Reading: Survey on Active Learning
Week 12 (November 8): Active Learning
Monday: Active Learning: Discriminative and Representative keynote, pdf
Assignment 4 Due
Required Reading: Active SVM for Text Classification, Active Learning using Pre-Clustering
Tuesday:
Assignment 5 Posted
Wednesday: Bayesian and Deep Active Learning keynote, pdf
Required Reading: Bayesian Active Learning, Deep Bayesian Active Learning
Week 13 (November 15): Zero-shot Learning
Monday: Zero-shot Learning: Overview keynote, pdf
Required Reading: Discriminative ZSL, Similarity-based ZSL, Embarrassingly Simple ZSL
Wednesday: Deep Zero-shot Learning keynote, pdf
Assignment 5 Due
Required Reading: Feature-generating networks, Aligned VAEs
Week 14 (November 22): Thanksgiving Break - no classes
Monday:
Wednesday:
Week 15 (November 29): Few-shot Learning, Meta Learning
Monday: Project Review and Feedback
Wednesday: Few-shot and meta-learning keynote, pdf
Required Reading: Matching Networks, Prototypical Networks, MAML
Week 16 (December 6): Project Presentations
Monday: Project Presentations