Course contents

  • Defining machine learning and data mining
  • Relation to other fields (stats, databases, probability, information theory)
  • Scalability
  • Privacy issues and social impact
  • Applications in AI, computer vision, computer games, search engines, marketing, bioinformatics, robotics, HCI and graphics.

    Exploratory Data Analysis
  • Linear algebra revision (eigenvectors !!!)
  • Pagerank
  • The SVD, spectral methods and latent semantic indexing
  • Probabilistic component analysis
  • Examples: text mining, search engines, image compression and visualization

    Graphical models
  • Introduction to discrete probability
  • Inference in Bayesian networks
  • Maximum likelihood and Bayesianlearning
  • Model selection

    Supervised learning
  • Introduction to continuous probability
  • Linear regression and classification (least squares and ridge)
  • Model assessment and cross-validation
  • Introduction to optimization
  • Nonlinear regression (neural nets and Gaussian processes)
  • Boosting and feature selection
  • Examples

    Unsupervised learning
  • Nearest neighbours and K-means
  • Spectral kernel methods for clustering and semi-supervised learning
  • The EM algorithm
  • Mixture models for discrete and continuous data
  • Temporal methods: hidden Markov models & Kalman filters
  • Boltzmann machines and random fields
  • Examples: web mining, collaborative filtering, music and image clustering, automatic translation, spam filtering, computer games and object recognition.

    Other forms of learning
  • Semi-supervised learning
  • Active learning
  • Reinforcement learning
  • Self-taught learning
  • LATEST :

    • The machine learning book of Hastie, Tibshirani and Friedman is now online: The elements of statistical learning.
    • Chapters 14,15 and 20 of the artificial intelligence book Stuart Russell and Peter Norvig is strongly recommended reading for this course. I'll provide partial photocopies of chapters 14 and 15 in class. Chapter 20 is available online.
    • This AIspace page at UBC has lots of videos and applets about inference in directed probabilistic graphical models (aka Bayesian networks or belief networks).
    • For graphical models and Beta-Bernoulli models, I recommend A Tutorial on Learning with Bayesian Networks David Heckerman.
    • Kevin Murphy has compiled a nice page about Bayesian learning.
    • Wikipedia tutorial on the: SVD
    • The following handout should help you with linear algebra revision: PDF
    • The homework should be handed in on Wednesday at the beginning of the class. Please note that messy homeworks will be penalized - it is your responsibility to ensure that the material is presented in a clear written form. All pseudocode must be handed in. Please don't forget to add your name and student number.