Hopfield Nets in Practice & Boltzmann machines

Q1: Can you give a example illustrate how we can use the hopfield nets in a learning task?

Q2: What’s the different between Boltzmann machine and the hopfield nets with stochastic units(Simulating annealling) ?


1 Response to “Hopfield Nets in Practice & Boltzmann machines”

  1. 1 Pierre Luc Carrier March 15, 2013 at 12:58

    Q1 :
    -Short answer : meh…
    -Long answer : Hopfield Nets are not really used anymore. One of the reason for this is that HN are trained to remember the training data exactly whereas the current main focus of machine learning is generalisation to data not seen at training time.

    Q2 :
    -In a Hopfield Net, all the units are visible (we can observe the state of all neurons) whereas a Boltzmann Machine has visible and hidden units (we can only observe the state of some neurons).
    -A Bolztmann Machine is trained with approximate likelihood.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: