Bidirectional associative memory

Bidirectional associative memory (BAM) is a type of recurrent neural network. BAM was introduced by Bart Kosko in 1988.^[1] There are two types of associative memory, auto-associative and hetero-associative. BAM is hetero-associative, meaning given a pattern it can return another pattern which is potentially of a different size. It is similar to the Hopfield network in that they are both forms of associative memory. However, Hopfield nets return patterns of the same size.

Contents 1 Topology 2 Procedure 2.1 Learning 2.2 Recall 3 See also 4 References 5 External links

Topology

It contains two layers of neurons one we shall call X and Y. Layer X and Y are fully connected with each other. Once the weights have been established, input into layer X presents the pattern in layer Y, and vice versa.

Procedure

Learning

Imagine we wish to store two associations, A1:B1 and A2:B2.

• A1 = (1, 0, 1, 0, 1, 0), B1 = (1, 1, 0, 0)
• A2 = (1, 1, 1, 0, 0, 0), B2 = (1, 0, 1, 0)

These are then transformed into the bipolar forms:

• X1 = (1, -1, 1, -1, 1, -1), Y1 = (1, 1, -1, -1)
• X2 = (1, 1, 1, -1, -1, -1), Y2 = (1, -1, 1, -1)

From there, we calculate where denotes the transpose. So,

Recall

To retrieve the association A1, we multiply it by M to get (4, 2, -2, -4), which, when run through a threshold, yields (1, 1, 0, 0), which is B1. To find the reverse association, multiply this by the transpose of M.

See also

• Autoassociative memory
• Self-organizing feature map

References

External links

• Bidirectional Associative Memory - online demo and perl source code
• Bidirectional associative memories - ACM Portal Reference