I'm struggling to understand the meaning of the term associative memory in the context of Hopfield networks. I'm not sure if associative learning is the same thing or a different thing. In associative learning, you have two unrelated items and you learn to associate one item with the other. In a hopfield network you have a partial pattern then you can retrieve the full pattern. How is this associative learning? Or are associative learning and associative memory different things?
$\begingroup$
$\endgroup$
2
-
$\begingroup$ Welcome to Psychology.SE Liyuan. In what context are you looking at associative memory? What have you been reading which lead you to this term? $\endgroup$– Chris RogersMay 12, 2018 at 17:18
-
$\begingroup$ From a computational neuroscience perspective, I'm studying network models of associative memory including hopfield network and the memory matrix. $\endgroup$– liyuanMay 12, 2018 at 17:35
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
An Associative Memory is a synonym for Content Addressable Memory or a memory which stores the result of associative learning, wherein an input item is learned to be associated with an output item.
A Hopfield network is a type of associative memory, wherein an input item (including partial, noisy representations of the original trained/encoded item) is an associated to a the ideal representation of the trained/desired/encoded input item.
For further discussion of Hopfield networks, I recommend look at the question "What's the difference between a Hamming and Hopfield network?"
answered May 12, 2018 at 20:36