Thresholded learning matrix for efficient pattern recalling

Mario Aldape-Pérez, Israel Román-Godínez, Oscar Camacho-Nieto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

The Lernmatrix, which is the first known model of associative memory, is a heteroassociative memory that can easily work as a binary pattern classifier if output patterns are appropriately chosen. However, this mathematical model undergoes fundamental patterns misclassification whenever crossbars saturation occurs. In this paper, a novel algorithm that overcomes Lernmatrix weaknesses is proposed. The crossbars saturation occurrence is solved by means of a dynamic threshold value which is computed for each recalled pattern. The algorithm applies the dynamic threshold value over the ambiguously recalled class vector in order to obtain a sentinel vector which is used for uncertainty elimination purposes. The efficiency and effectiveness of our approach is demonstrated through comparisons with other methods using real-world data. © 2008 Springer-Verlag Berlin Heidelberg.
Original languageAmerican English
Title of host publicationThresholded learning matrix for efficient pattern recalling
Pages445-452
Number of pages8
ISBN (Electronic)3540859195
DOIs
StatePublished - 10 Nov 2008
EventLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) -
Duration: 10 Nov 2008 → …

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5197 LNCS
ISSN (Print)0302-9743

Conference

ConferenceLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Period10/11/08 → …

Fingerprint

Data storage equipment
Threshold Value
Saturation
Classifiers
Mathematical models
Misclassification
Associative Memory
Elimination
Classifier
Mathematical Model
Binary
Uncertainty
Learning
Output
Model
Class

Cite this

Aldape-Pérez, M., Román-Godínez, I., & Camacho-Nieto, O. (2008). Thresholded learning matrix for efficient pattern recalling. In Thresholded learning matrix for efficient pattern recalling (pp. 445-452). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5197 LNCS). https://doi.org/10.1007/978-3-540-85920-8_55
Aldape-Pérez, Mario ; Román-Godínez, Israel ; Camacho-Nieto, Oscar. / Thresholded learning matrix for efficient pattern recalling. Thresholded learning matrix for efficient pattern recalling. 2008. pp. 445-452 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{7af668673e51451cb1be9ed03180c5c2,
title = "Thresholded learning matrix for efficient pattern recalling",
abstract = "The Lernmatrix, which is the first known model of associative memory, is a heteroassociative memory that can easily work as a binary pattern classifier if output patterns are appropriately chosen. However, this mathematical model undergoes fundamental patterns misclassification whenever crossbars saturation occurs. In this paper, a novel algorithm that overcomes Lernmatrix weaknesses is proposed. The crossbars saturation occurrence is solved by means of a dynamic threshold value which is computed for each recalled pattern. The algorithm applies the dynamic threshold value over the ambiguously recalled class vector in order to obtain a sentinel vector which is used for uncertainty elimination purposes. The efficiency and effectiveness of our approach is demonstrated through comparisons with other methods using real-world data. {\circledC} 2008 Springer-Verlag Berlin Heidelberg.",
author = "Mario Aldape-P{\'e}rez and Israel Rom{\'a}n-God{\'i}nez and Oscar Camacho-Nieto",
year = "2008",
month = "11",
day = "10",
doi = "10.1007/978-3-540-85920-8_55",
language = "American English",
isbn = "3540859195",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "445--452",
booktitle = "Thresholded learning matrix for efficient pattern recalling",

}

Aldape-Pérez, M, Román-Godínez, I & Camacho-Nieto, O 2008, Thresholded learning matrix for efficient pattern recalling. in Thresholded learning matrix for efficient pattern recalling. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5197 LNCS, pp. 445-452, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 10/11/08. https://doi.org/10.1007/978-3-540-85920-8_55

Thresholded learning matrix for efficient pattern recalling. / Aldape-Pérez, Mario; Román-Godínez, Israel; Camacho-Nieto, Oscar.

Thresholded learning matrix for efficient pattern recalling. 2008. p. 445-452 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5197 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Thresholded learning matrix for efficient pattern recalling

AU - Aldape-Pérez, Mario

AU - Román-Godínez, Israel

AU - Camacho-Nieto, Oscar

PY - 2008/11/10

Y1 - 2008/11/10

N2 - The Lernmatrix, which is the first known model of associative memory, is a heteroassociative memory that can easily work as a binary pattern classifier if output patterns are appropriately chosen. However, this mathematical model undergoes fundamental patterns misclassification whenever crossbars saturation occurs. In this paper, a novel algorithm that overcomes Lernmatrix weaknesses is proposed. The crossbars saturation occurrence is solved by means of a dynamic threshold value which is computed for each recalled pattern. The algorithm applies the dynamic threshold value over the ambiguously recalled class vector in order to obtain a sentinel vector which is used for uncertainty elimination purposes. The efficiency and effectiveness of our approach is demonstrated through comparisons with other methods using real-world data. © 2008 Springer-Verlag Berlin Heidelberg.

AB - The Lernmatrix, which is the first known model of associative memory, is a heteroassociative memory that can easily work as a binary pattern classifier if output patterns are appropriately chosen. However, this mathematical model undergoes fundamental patterns misclassification whenever crossbars saturation occurs. In this paper, a novel algorithm that overcomes Lernmatrix weaknesses is proposed. The crossbars saturation occurrence is solved by means of a dynamic threshold value which is computed for each recalled pattern. The algorithm applies the dynamic threshold value over the ambiguously recalled class vector in order to obtain a sentinel vector which is used for uncertainty elimination purposes. The efficiency and effectiveness of our approach is demonstrated through comparisons with other methods using real-world data. © 2008 Springer-Verlag Berlin Heidelberg.

U2 - 10.1007/978-3-540-85920-8_55

DO - 10.1007/978-3-540-85920-8_55

M3 - Conference contribution

SN - 3540859195

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 445

EP - 452

BT - Thresholded learning matrix for efficient pattern recalling

ER -

Aldape-Pérez M, Román-Godínez I, Camacho-Nieto O. Thresholded learning matrix for efficient pattern recalling. In Thresholded learning matrix for efficient pattern recalling. 2008. p. 445-452. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-85920-8_55