An Echo State Neural Network for Foetal Electrocardiogram Extraction Optimised by Random Search

Joachim Behar; Alistair E. W. Johnson; Julien Oster; Garid D. Clifford

An Echo State Neural Network for Foetal Electrocardiogram Extraction Optimised by Random Search

Joachim Behar, Alistair E. W. Johnson, Julien Oster, Garid D. Clifford

Research output: Contribution to conference › Paper › peer-review

30 Downloads (Pure)

Abstract

We present a novel application of an echo state neural network (ESN) to noninvasive foetal electrocardiogram (FECG) extraction. Extraction of the FECG is performed on abdominal recordings of pregnant women via maternal ECG cancellation. The FECG can then be used for foetal health monitoring by extracting clinically interpretable features. We show that optimising an ESN by random search gives almost equivalent performance to an exhaustive grid search with 85.6% vs. 87.9% accuracy on the test database. This is particularly useful as, while powerful, ESNs have many hyper-parameters which are not easily optimised using expert knowledge.

Original language	American English
State	Published - 2013
Externally published	Yes
Event	27th Annual Conference on Neural Information Processing Systems, NIPS 2013 - Lake Tahoe, NV, United States Duration: 5 Dec 2013 → 10 Dec 2013

Conference

Conference	27th Annual Conference on Neural Information Processing Systems, NIPS 2013
Country/Territory	United States
City	Lake Tahoe, NV
Period	5/12/13 → 10/12/13

Access to Document

BeharJohnsonOsterClifford_RS_example

Cite this

@conference{fdc1ffbf7b9c4cbb94402280e00e88ef,

title = "An Echo State Neural Network for Foetal Electrocardiogram Extraction Optimised by Random Search",

abstract = "We present a novel application of an echo state neural network (ESN) to noninvasive foetal electrocardiogram (FECG) extraction. Extraction of the FECG is performed on abdominal recordings of pregnant women via maternal ECG cancellation. The FECG can then be used for foetal health monitoring by extracting clinically interpretable features. We show that optimising an ESN by random search gives almost equivalent performance to an exhaustive grid search with 85.6% vs. 87.9% accuracy on the test database. This is particularly useful as, while powerful, ESNs have many hyper-parameters which are not easily optimised using expert knowledge.",

author = "Joachim Behar and Johnson, {Alistair E. W.} and Julien Oster and Clifford, {Garid D.}",

year = "2013",

language = "American English",

note = "27th Annual Conference on Neural Information Processing Systems, NIPS 2013 ; Conference date: 05-12-2013 Through 10-12-2013",

}

TY - CONF

T1 - An Echo State Neural Network for Foetal Electrocardiogram Extraction Optimised by Random Search

AU - Behar, Joachim

AU - Johnson, Alistair E. W.

AU - Oster, Julien

AU - Clifford, Garid D.

PY - 2013

Y1 - 2013

N2 - We present a novel application of an echo state neural network (ESN) to noninvasive foetal electrocardiogram (FECG) extraction. Extraction of the FECG is performed on abdominal recordings of pregnant women via maternal ECG cancellation. The FECG can then be used for foetal health monitoring by extracting clinically interpretable features. We show that optimising an ESN by random search gives almost equivalent performance to an exhaustive grid search with 85.6% vs. 87.9% accuracy on the test database. This is particularly useful as, while powerful, ESNs have many hyper-parameters which are not easily optimised using expert knowledge.

AB - We present a novel application of an echo state neural network (ESN) to noninvasive foetal electrocardiogram (FECG) extraction. Extraction of the FECG is performed on abdominal recordings of pregnant women via maternal ECG cancellation. The FECG can then be used for foetal health monitoring by extracting clinically interpretable features. We show that optimising an ESN by random search gives almost equivalent performance to an exhaustive grid search with 85.6% vs. 87.9% accuracy on the test database. This is particularly useful as, while powerful, ESNs have many hyper-parameters which are not easily optimised using expert knowledge.

UR - https://nips.cc/Conferences/2013/Schedule?showEvent=3715

M3 - Paper

T2 - 27th Annual Conference on Neural Information Processing Systems, NIPS 2013

Y2 - 5 December 2013 through 10 December 2013

ER -

An Echo State Neural Network for Foetal Electrocardiogram Extraction Optimised by Random Search

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this