I've been trying to implement the Boltzmann machine 4-2-4 encoder that appeared in A Learning Algorithm for Boltzmann Machines. but I am unable to find clear pseudocode for doing it or more specific details. I have tried "Neural networks: algorithms, applications, and programming techniques" and Fausett books so far. Where can I find an implementation of the encoder or more specific details about it?


I finally implemented the Boltzmann machine for 4-2-4 encoder problem using several descriptions. You can find my implementation in this blog post.

  • $\begingroup$ Just a note: the $\LaTeX$ code on your page isn't rendering on my computer, it's just source code. Not sure if it's because I'm using an old browser on an old computer or what. $\endgroup$ – shadowtalker Jun 25 '14 at 3:17
  • $\begingroup$ It is a problem of loading correctly the MathJax libraries on blogspot. It seems that I managed to solve the Latex rendering issue using the advice on this page link, hope it works for you. $\endgroup$ – Gorayni Jun 25 '14 at 4:21
  • $\begingroup$ Still no luck, but MathJax seems to have stopped working on other sites as well for me so I don't think you can fix it, unless you want to buy me a new laptop. $\endgroup$ – shadowtalker Jun 25 '14 at 4:59

Salakhutdinov was a student of Hinton who has since moved to Microsoft.

http://www.cs.toronto.edu/~rsalakhu/DBM.html <-- here is working MatLab code for an RBM.

It's a bit of a mess, I'm a week into rewriting it. If I haven't updated this answer within a month, someone ping me and I will upload my rewrite to GitHub.


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.