These are the lectures for a graduate machine learning course taught by Nando de Freitas at UBC in 2013.
Professor: Nando de Freitas
University of British Columbia
Date: January, 2013
Th Jan 17. Ridge, nonlinear regression with basis functions and Cross-validation.
Th Jan 24. Bayesian learning (part I).
Th Jan 31. Gaussian processes for nonlinear regression (part I).
Tue Feb 5. Gaussian processes for nonlinear regression (part II). Python demo code for GP regression.
Tue Feb 19. Spring break.
Th Feb 21. Spring break.
Tue Feb 26. Random forests applications: Object detection and Kinect.
Th Mar 14. Deep learning with autoencoders II.
Tue Mar 19. Importance sampling and MCMC.
Th Mar 21. Importance sampling and MCMC.
Th Mar 28. Constrained optimization: Lagrangians and duality. Application to penalized maximum likelihood and Lasso. (notes from 2011 course – the lecture will actually be used to ask project questions.)
Tue Apr 2. Exam Revision.
Th Apr 4. Exam.
Tue Apr 16. Project due.
Tue Apr 23. Reviews due.