Scribd
Upload
75 views

Associative Memory

Associative memory networks are inspired by the associative memory function of the brain. They allow for content-addressable memory retrieval using partial or degraded key patterns. There are three main types: auto-associative which associates patterns with themselves, hetero-associative which associates different patterns, and bidirectional which allows associations in both directions. Training algorithms like Hebbian learning and the outer product rule are used to encode associative memories by updating the connection weights between neurons. Encoding stores the patterns while decoding retrieves them based on input keys.

Uploaded by

rashi
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
75 views

Associative Memory

Associative memory networks are inspired by the associative memory function of the brain. They allow for content-addressable memory retrieval using partial or degraded key patterns. There are three main types: auto-associative which associates patterns with themselves, hetero-associative which associates different patterns, and bidirectional which allows associations in both directions. Training algorithms like Hebbian learning and the outer product rule are used to encode associative memories by updating the connection weights between neurons. Encoding stores the patterns while decoding retrieves them based on input keys.

Uploaded by

rashi
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 15

ASSOCIATIVE MEMORY

NETWORKS

.
ASSOCIATIVE MEMORY
NETWORK
• One of the primary functions of the brain is
associative memory. We associate the faces
with names, letters with sounds, or we can
recognize the people even if they have
sunglasses or if they are somehow elder now.
• Recall through Key pattern
• Content addressable memory (not address
addressable memory)

2
April 2007 3
Types of Associative memories:-
(1) Auto-associative memory
(2) Hetro-associative memory
(3) Bi-directional memory

April 2007 4
(1)Auto-associative memory

5
(2)Hetero-associative memory

6
(3)Bidirectional associative
memory
BAM is also Known as Recurrent hetero-associative
matching network

April 2007 7
Start

Initialize weights to Zero

For i=1 to n
For j=1 to n
For
each
s:t

Flowchart
For
Training
Process of
Autoassoc-
Continue
iatve
Continue network

Stop
Training ALGORITHM:-

P
W ( p)   s T ( p)  t ( p)
p 1
Testing ALGORITHM:-
TRAINING ALGORITHM FOR Pattern
Association)

There are two Algorithms developed for training of Pattern association


nets.
(1) Hebb Rule
(2) Outer product rule
(1)HEBB RULE FOR PATTERN ASSOCIATION
 Hebb Rule already Done( Refer Previous Slides)
(2) OUTER PRODUCT FOR PATTERN
ASSOCIATION
Let s and t be row vectors.

Then for a particular training pair s:t

Input  s
Output  t
W ( p )  s T ( p )  t ( p ) 
 s1   s1t1...s1t m   w11 ...w1m 
     
   t ,..., t    s2t1...s2t m    
  1 m
   
     
 sn   sn t1...sn t m   wn1 ...wnm 

and
P
W ( p)   s T ( p)  t ( p)
p 1
• Encoding:-The process of constructing the
connection weigh matrix is called encoding
• Decoding:-The process of retrieving a stored
pattern given an input pattern is called
decoding
• Storage capacity:-Number of patterns that
can be correctly stored & recalled by a
network.

14
Reference:-

BOOK:-
“Principles of Soft Computing, 2nd
Edition”
by S.N. Sivanandam & SN Deepa

15

You might also like