Are there any good books for me to learn probability programming? For example, I am new to Latent Dirichlet allocation (LDA) and Gibbs sampling. I have read some books about the techniques, but it still feels hard to develop programs to apply them. Are there any books that can help?

I have taken some courses in machine learning, statistics and data mining. I am familiar with discriminant classifiers such as Support Vector Machines and the basic knowledge in generative classification. So I understand the concept in Naive Bayes Classification and Language models in Information Retrieval. I am new to the field of Latent variable analysis and graph modeling. Specifically, are there any books about how to develop programs related sampling, bayes inference? For instance, if I want to implement LDA using Gibbs sampling, any book can help? It is preferable if it has some easy understandable codes using toy examples.

  • 2
    $\begingroup$ Your question is rather broad - it's hard to identify exactly what kind of assistance you need and therefore hard to make any suggestions. Could you clarify your needs? $\endgroup$
    – Glen_b
    Aug 6 '13 at 10:01
  • 3
    $\begingroup$ D in LDA = discriminant or Dirichlet? That's a separator in itself. Whether you think you are doing statistics or machine learning will have some influence on what you most need or will find most congenial. What software do you use or intend to use? $\endgroup$
    – Nick Cox
    Aug 6 '13 at 10:04
  • 1
    $\begingroup$ It would also help if you could mention the books you've read and your background. $\endgroup$
    – QuantIbex
    Aug 6 '13 at 10:42
  • 1
    $\begingroup$ I edited the question. I have taken some courses in Machine Learning, Data Mining and Statistics. Thanks! $\endgroup$
    – user28859
    Aug 6 '13 at 19:17

See Probabilistic Programming and Bayesian Methods for Hackers. It is an interactive book that explores probabilistic programming in Python.


The BUGS book is pretty decent (if you want to learn BUGS).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.