Some suggestions for projects

  1. Restricted Boltzmann machines (, Murphy pg 987)
  2. Loopy belief propagation with weights
  3. Solving K-SAT using survey propagation (
  4. Adaptive belief propagation (updating only subsets of variables at a time)
  5. Ranking players using Trueskill (Murphy, page 795 and references)
  6. Tree-reweighted Belief propagation (Wainwright, Jaakola, Wilsky, "A New Class of Upper bound on the Log-Partition Function")
  7. Gaussian belief propagation and solutions to linear equations (
  8. Compressed sensing
    1. Approximate message passing (Papers by Montanari, Donoho, Rangan, Schniter)
    2. Compressed sensing algorithms and matrices based on sparse graphs (Lee, Pedarsani, Ramchandran and Vem, Janakiraman and Narayanan)
    3. Group testing, network traffic analysis
  9. Stochastic block models, message passing algorithms for community detection (
  10. Monte Carlo methods
    1. Particle filtering (
    2. Markov Chain Monte Carlo - Gibbs sampling, Metropolis Hastings algorithm (Mackay's book, Murphy's book pg. 839)
  11. Learning graphical models using graph lasso (Start with Murphy, pg 944, Tibshirani et al)
  12. Learning parameters of a state-space model using the Baum-Welch algorithm (Moon's book on Mathematical methods in signal processing, papers)
  13. Structure learning (Chow-Liu algorithm)
  14. Applications
    1. Hidden Markov Models for speech processing
    2. Markov models for text recognition
    3. Collaborative filtering
    4. Ranking algorithms (Trueskill)
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License