1. Jones and Hobert (2001). "Honest Exploration of Intractable Probability Distributions via Markov Chain Monte Carlo," Statistical Science 16 (4), 312--334 [PDF]
  2. Wei and Tanner (1990). "A Monte Carlo implementation of the EM algorithm and the poor man's data augmentation algorithms," JASA, 85 (411) 699--704 [PDF]
    • How does the MCEM algorithm differ from the EM algorithm? What's the basic principle used?
    • According to Wei and Tanner, what is the difference between the EM algorithm and an "EM-type" algorithm?
    • When might the MCEM algorithm not work well (or at least be difficult to implement)?
    • What are the advantages and disadvantages of using the "poor man's" data augmentation algorithm? Which algorithm (MCEM or PMDA) is more computationally intense?
    • When is the "poor man's" data augmentation algorithm likely to be close to results obtained using the true data augmentation algorithm?
  4. Coding standards for Biostat 778
  5. Appendix A and B from Tanenbaum
  6. Tanenbaum-Tovalds Debate