Abstract
This paper proposes a method to compute exact amount of noise in recalling every encoded pattern. This can be used to find vectors for construction of augmented matrices for both Wang et al., and Wang-Lee methods. The computation of noise also gives a method for obtaining minimum dimension of augmented vectors. This information is used to construct augmented matrix for guaranteed recall of every encoded pattern. A three layer Bidirectional Associative Memory (TLBAM) that requires smaller size correlation matrices is also proposed. The number of elements in the correlation matrices are proportional to the number of pattern-pairs encoded. The TLBAM when encoded using the novel augmentation method guarantees correct recall of all stored pairs. The encoded network will also require much fewer elements than any other method.
| Original language | English (US) |
|---|---|
| State | Published - Jan 1 1996 |
| Event | Proceedings of the 1996 IEEE International Conference on Neural Networks, ICNN. Part 1 (of 4) - Washington, DC, USA Duration: Jun 3 1996 → Jun 6 1996 |
Other
| Other | Proceedings of the 1996 IEEE International Conference on Neural Networks, ICNN. Part 1 (of 4) |
|---|---|
| City | Washington, DC, USA |
| Period | 6/3/96 → 6/6/96 |
Fingerprint
ASJC Scopus subject areas
- Software
Cite this
Three-stage architecture for bidirectional associative memory. / Sarkar, Dilip.
1996. Paper presented at Proceedings of the 1996 IEEE International Conference on Neural Networks, ICNN. Part 1 (of 4), Washington, DC, USA, .Research output: Contribution to conference › Paper
}
TY - CONF
T1 - Three-stage architecture for bidirectional associative memory
AU - Sarkar, Dilip
PY - 1996/1/1
Y1 - 1996/1/1
N2 - This paper proposes a method to compute exact amount of noise in recalling every encoded pattern. This can be used to find vectors for construction of augmented matrices for both Wang et al., and Wang-Lee methods. The computation of noise also gives a method for obtaining minimum dimension of augmented vectors. This information is used to construct augmented matrix for guaranteed recall of every encoded pattern. A three layer Bidirectional Associative Memory (TLBAM) that requires smaller size correlation matrices is also proposed. The number of elements in the correlation matrices are proportional to the number of pattern-pairs encoded. The TLBAM when encoded using the novel augmentation method guarantees correct recall of all stored pairs. The encoded network will also require much fewer elements than any other method.
AB - This paper proposes a method to compute exact amount of noise in recalling every encoded pattern. This can be used to find vectors for construction of augmented matrices for both Wang et al., and Wang-Lee methods. The computation of noise also gives a method for obtaining minimum dimension of augmented vectors. This information is used to construct augmented matrix for guaranteed recall of every encoded pattern. A three layer Bidirectional Associative Memory (TLBAM) that requires smaller size correlation matrices is also proposed. The number of elements in the correlation matrices are proportional to the number of pattern-pairs encoded. The TLBAM when encoded using the novel augmentation method guarantees correct recall of all stored pairs. The encoded network will also require much fewer elements than any other method.
UR - http://www.scopus.com/inward/record.url?scp=0029767527&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0029767527&partnerID=8YFLogxK
M3 - Paper
ER -