An associative memory for the on-line recognition and prediction of temporal sequences

J. Bose; S. B. Furber; J. L. Shapiro

doi:10.1109/IJCNN.2005.1556028

An associative memory for the on-line recognition and prediction of temporal sequences

J. Bose, S. B. Furber, J. L. Shapiro

Advanced Processor Technology

Research output: Chapter in Book/Conference proceeding › Chapter › peer-review

Abstract

This paper presents the design of an associative memory with feedback that is capable of on-line temporal sequence learning. A framework for on-line sequence learning has been proposed, and different sequence learning models have been analysed according to this framework. The network model is an associative memory with a separate store for the sequence context of a symbol. A sparse distributed memory is used to gain scalability. The context store combines the functionality of a neural layer with a shift register. The sensitivity of the machine to the sequence context is controllable, resulting in different characteristic behaviours. The model can store and predict online sequences of various types and length. Numerical simulations on the model have been carried out to determine its properties. © 2005 IEEE.

Original language	English
Title of host publication	Proceedings of the International Joint Conference on Neural Networks (IJCNN) 2005 : July 31 - August 4, 2005, Hilton Montréal Bonaventure Hotel, Montréal, Québec, Canada
Publisher	IEEE
Pages	1223-1228
Number of pages	5
Volume	2
ISBN (Print)	0-7803-9048-2
DOIs	https://doi.org/10.1109/IJCNN.2005.1556028
Publication status	Published - 2005

Access to Document

10.1109/IJCNN.2005.1556028

Cite this

Bose, J., Furber, S. B., & Shapiro, J. L. (2005). An associative memory for the on-line recognition and prediction of temporal sequences. In Proceedings of the International Joint Conference on Neural Networks (IJCNN) 2005 : July 31 - August 4, 2005, Hilton Montréal Bonaventure Hotel, Montréal, Québec, Canada (Vol. 2, pp. 1223-1228). IEEE. https://doi.org/10.1109/IJCNN.2005.1556028

@inbook{5457a29341e7448488c6913a68b6d09e,

title = "An associative memory for the on-line recognition and prediction of temporal sequences",

abstract = "This paper presents the design of an associative memory with feedback that is capable of on-line temporal sequence learning. A framework for on-line sequence learning has been proposed, and different sequence learning models have been analysed according to this framework. The network model is an associative memory with a separate store for the sequence context of a symbol. A sparse distributed memory is used to gain scalability. The context store combines the functionality of a neural layer with a shift register. The sensitivity of the machine to the sequence context is controllable, resulting in different characteristic behaviours. The model can store and predict online sequences of various types and length. Numerical simulations on the model have been carried out to determine its properties. {\textcopyright} 2005 IEEE.",

author = "J. Bose and Furber, {S. B.} and Shapiro, {J. L.}",

note = "informal publication",

year = "2005",

doi = "10.1109/IJCNN.2005.1556028",

language = "English",

isbn = "0-7803-9048-2",

volume = "2",

pages = "1223--1228",

booktitle = "Proceedings of the International Joint Conference on Neural Networks (IJCNN) 2005 : July 31 - August 4, 2005, Hilton Montr{\'e}al Bonaventure Hotel, Montr{\'e}al, Qu{\'e}bec, Canada",

publisher = "IEEE",

address = "United States",

}

Bose, J, Furber, SB & Shapiro, JL 2005, An associative memory for the on-line recognition and prediction of temporal sequences. in Proceedings of the International Joint Conference on Neural Networks (IJCNN) 2005 : July 31 - August 4, 2005, Hilton Montréal Bonaventure Hotel, Montréal, Québec, Canada. vol. 2, IEEE, pp. 1223-1228. https://doi.org/10.1109/IJCNN.2005.1556028

An associative memory for the on-line recognition and prediction of temporal sequences. / Bose, J.; Furber, S. B.; Shapiro, J. L.
Proceedings of the International Joint Conference on Neural Networks (IJCNN) 2005 : July 31 - August 4, 2005, Hilton Montréal Bonaventure Hotel, Montréal, Québec, Canada. Vol. 2 IEEE, 2005. p. 1223-1228.

Research output: Chapter in Book/Conference proceeding › Chapter › peer-review

TY - CHAP

T1 - An associative memory for the on-line recognition and prediction of temporal sequences

AU - Bose, J.

AU - Furber, S. B.

AU - Shapiro, J. L.

N1 - informal publication

PY - 2005

Y1 - 2005

N2 - This paper presents the design of an associative memory with feedback that is capable of on-line temporal sequence learning. A framework for on-line sequence learning has been proposed, and different sequence learning models have been analysed according to this framework. The network model is an associative memory with a separate store for the sequence context of a symbol. A sparse distributed memory is used to gain scalability. The context store combines the functionality of a neural layer with a shift register. The sensitivity of the machine to the sequence context is controllable, resulting in different characteristic behaviours. The model can store and predict online sequences of various types and length. Numerical simulations on the model have been carried out to determine its properties. © 2005 IEEE.

AB - This paper presents the design of an associative memory with feedback that is capable of on-line temporal sequence learning. A framework for on-line sequence learning has been proposed, and different sequence learning models have been analysed according to this framework. The network model is an associative memory with a separate store for the sequence context of a symbol. A sparse distributed memory is used to gain scalability. The context store combines the functionality of a neural layer with a shift register. The sensitivity of the machine to the sequence context is controllable, resulting in different characteristic behaviours. The model can store and predict online sequences of various types and length. Numerical simulations on the model have been carried out to determine its properties. © 2005 IEEE.

U2 - 10.1109/IJCNN.2005.1556028

DO - 10.1109/IJCNN.2005.1556028

M3 - Chapter

SN - 0-7803-9048-2

VL - 2

SP - 1223

EP - 1228

BT - Proceedings of the International Joint Conference on Neural Networks (IJCNN) 2005 : July 31 - August 4, 2005, Hilton Montréal Bonaventure Hotel, Montréal, Québec, Canada

PB - IEEE

ER -

An associative memory for the on-line recognition and prediction of temporal sequences

Abstract

Access to Document

Fingerprint

Cite this