This course is intended as a second or third university-level course on machine learning, a field that focuses on using automated data analysis for tasks like pattern recognition and prediction. The class is intended as a continuation of CPSC 340 (or 532M), and will assume a strong background in math and computer science. Topics will (roughly) include deep learning, generative models, latent-variable models, Markov models, probabilistic graphical models, and Bayesian methods.
- Least squares using Gaussian radial basis functions (RBFs) and L2-regularization
- Multi-class logistic regression with a softmax classifier
- Principal component analysis
- Minimizing strictly-convex quadratic functions
- MAP estimation
- Machine learning model memory and time complexities
- Gradients and hessians in matrix notation
- Proving norm inequalities
- Convexity
- MLE and MAP estimation for general discrete distribution
- Gaussian self-conjugacy
- Generative classifiers with Gaussian assumption
- Expectation maximization for categorical mixture models
- Semi-supervised Gaussian discriminant analysis
- Approximate and exact inference in discrete Markov chains
- Viterbi decoding algorithm for Markov chains
- Inference in continuous-state Markov chains
- Learning with homogeneous Markov chains
- D-separation and exact inference in DAGs
- Learning and sample generation with DAGs
- Inference in undirected graphical models
- Conditional undirected graphical models
- Lattice-structured conditional UGM
- Conjugate priors
- Empirical Bayes
- Metropolis-coupled Markov chain Monte Carlo