Noise Performance of Linear Associative Memories

Kalavai J. Raghunath; Vladimir Cherkassky

doi:10.1109/34.297959

Noise Performance of Linear Associative Memories

Kalavai J. Raghunath, Vladimir Cherkassky

Electrical and Computer Engineering

Research output: Contribution to journal › Article › peer-review

3 Scopus citations

Abstract

The performance of two commonly used linear models of associative memories, generalized inverse (G1) and correlation matrix memory (CMM) is studied analytically in the presence of a new type of noise (training noise due to noisy training patterns). Theoretical expressions are determined for the SNR (signal-to-noise ratio) gain of the GI and CMM memories in the auto-associative and hetero-associative modes of operation. It is found that the GI method performance degrades significantly in the presence of training noise while the CMM method is relatively unaffected by it. The theoretical expressions are plotted and compared with the results obtained from Monte Carlo simulations and the two are found to be in excellent agreement.

Original language	English (US)
Pages (from-to)	757-765
Number of pages	9
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	16
Issue number	7
DOIs	https://doi.org/10.1109/34.297959
State	Published - Jul 1994

Keywords

Associative memory
correlation memory matrix
generalized inverse
noise performance

Publisher link

10.1109/34.297959

Find It

Find It at U of M Libraries

Cite this

@article{cd7e0b018e0344aca5298f08d737bdfa,

title = "Noise Performance of Linear Associative Memories",

abstract = "The performance of two commonly used linear models of associative memories, generalized inverse (G1) and correlation matrix memory (CMM) is studied analytically in the presence of a new type of noise (training noise due to noisy training patterns). Theoretical expressions are determined for the SNR (signal-to-noise ratio) gain of the GI and CMM memories in the auto-associative and hetero-associative modes of operation. It is found that the GI method performance degrades significantly in the presence of training noise while the CMM method is relatively unaffected by it. The theoretical expressions are plotted and compared with the results obtained from Monte Carlo simulations and the two are found to be in excellent agreement.",

keywords = "Associative memory, correlation memory matrix, generalized inverse, noise performance",

author = "Raghunath, {Kalavai J.} and Vladimir Cherkassky",

year = "1994",

month = jul,

doi = "10.1109/34.297959",

language = "English (US)",

volume = "16",

pages = "757--765",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "IEEE Computer Society",

number = "7",

}

TY - JOUR

T1 - Noise Performance of Linear Associative Memories

AU - Raghunath, Kalavai J.

AU - Cherkassky, Vladimir

PY - 1994/7

Y1 - 1994/7

N2 - The performance of two commonly used linear models of associative memories, generalized inverse (G1) and correlation matrix memory (CMM) is studied analytically in the presence of a new type of noise (training noise due to noisy training patterns). Theoretical expressions are determined for the SNR (signal-to-noise ratio) gain of the GI and CMM memories in the auto-associative and hetero-associative modes of operation. It is found that the GI method performance degrades significantly in the presence of training noise while the CMM method is relatively unaffected by it. The theoretical expressions are plotted and compared with the results obtained from Monte Carlo simulations and the two are found to be in excellent agreement.

AB - The performance of two commonly used linear models of associative memories, generalized inverse (G1) and correlation matrix memory (CMM) is studied analytically in the presence of a new type of noise (training noise due to noisy training patterns). Theoretical expressions are determined for the SNR (signal-to-noise ratio) gain of the GI and CMM memories in the auto-associative and hetero-associative modes of operation. It is found that the GI method performance degrades significantly in the presence of training noise while the CMM method is relatively unaffected by it. The theoretical expressions are plotted and compared with the results obtained from Monte Carlo simulations and the two are found to be in excellent agreement.

KW - Associative memory

KW - correlation memory matrix

KW - generalized inverse

KW - noise performance

UR - http://www.scopus.com/inward/record.url?scp=0028464564&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028464564&partnerID=8YFLogxK

U2 - 10.1109/34.297959

DO - 10.1109/34.297959

M3 - Article

AN - SCOPUS:0028464564

SN - 0162-8828

VL - 16

SP - 757

EP - 765

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 7

ER -

Noise Performance of Linear Associative Memories

Abstract

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this