The moving targets algorithm for difficult temporal credit assignment problems

Richard Rohwer; Fionn Murtagh

The moving targets algorithm for difficult temporal credit assignment problems

Richard Rohwer, Fionn Murtagh (Editor)

Research output: Unpublished contribution to conference › Unpublished Conference Paper › peer-review

Abstract

The 'moving targets' algorithm for training recurrent networks is reviewed and applied to a task which demonstrates the ability of this algorithm to use distant contextual information. Some practical difficulties are discussed, especially with regard to the minimization process. Results on performance and computational requirements of several different 2nd-order minimization algorithms are presented for moving target problems.

Original language	English
Publication status	Published - 1990
Event	Neural Networks for Statistical and Economic Data - Dublin, Republic of Ireland Duration: 10 Dec 1990 → 11 Dec 1990

Workshop

Workshop	Neural Networks for Statistical and Economic Data
City	Dublin, Republic of Ireland
Period	10/12/90 → 11/12/90

Keywords

networks
algorithm
minimization
moving target

Access to Document

misc91-007.pdf

Cite this

@conference{5b166dab728a42f2b34ff71622697dc1,

title = "The moving targets algorithm for difficult temporal credit assignment problems",

abstract = "The 'moving targets' algorithm for training recurrent networks is reviewed and applied to a task which demonstrates the ability of this algorithm to use distant contextual information. Some practical difficulties are discussed, especially with regard to the minimization process. Results on performance and computational requirements of several different 2nd-order minimization algorithms are presented for moving target problems.",

keywords = "networks, algorithm, minimization, moving target",

author = "Richard Rohwer and Fionn Murtagh",

year = "1990",

language = "English",

note = "Neural Networks for Statistical and Economic Data ; Conference date: 10-12-1990 Through 11-12-1990",

}

TY - CONF

T1 - The moving targets algorithm for difficult temporal credit assignment problems

AU - Rohwer, Richard

A2 - Murtagh, Fionn

PY - 1990

Y1 - 1990

N2 - The 'moving targets' algorithm for training recurrent networks is reviewed and applied to a task which demonstrates the ability of this algorithm to use distant contextual information. Some practical difficulties are discussed, especially with regard to the minimization process. Results on performance and computational requirements of several different 2nd-order minimization algorithms are presented for moving target problems.

AB - The 'moving targets' algorithm for training recurrent networks is reviewed and applied to a task which demonstrates the ability of this algorithm to use distant contextual information. Some practical difficulties are discussed, especially with regard to the minimization process. Results on performance and computational requirements of several different 2nd-order minimization algorithms are presented for moving target problems.

KW - networks

KW - algorithm

KW - minimization

KW - moving target

M3 - Unpublished Conference Paper

T2 - Neural Networks for Statistical and Economic Data

Y2 - 10 December 1990 through 11 December 1990

ER -