Classifying ischemic events using a bayesian inference multilayer perceptron and input variable evaluation using automatic relevance determination

M. G. Smyrnakis, D. J. Evans

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

Abstract

In this paper we present a Bayesian inference Multilayer Perceptron (MLP) which was used to classify the events of the Long Term ST Database (LTSTDB) as ischaemic or non-ischaemic episodes with an accuracy of 89.1%, sensitivity of 82.3% and specificity of 91.2% when the accuracy of the winning paper was 90.7%. The Automatic Relevance Determination (ARD) method was used to identify which of the extracted features that were used as input in the Bayesian inference MLP were the most important with respect to the models performance. ARD indicated that AT, a combination of the ST deviation and the duration of the episode, inspired from Langley et al [1], was the most important feature for determining Ischaemic episodes, given the data. A simple MLP which had as input variable of only AT was trained to verify the results of the ARD method. The classification accuracy was 85.8% on the test set. We can conclude from the results that the most important extracted feature was ΔT.

Original languageEnglish
Title of host publicationComputers in Cardiology
Pages305-308
Number of pages4
Volume34
DOIs
Publication statusPublished - 1 Dec 2007
EventComputers in Cardiology 2007, CAR 2007 - Durham, NC, United Kingdom
Duration: 30 Sep 20073 Oct 2007

Conference

ConferenceComputers in Cardiology 2007, CAR 2007
CountryUnited Kingdom
CityDurham, NC
Period30/09/073/10/07

Fingerprint

Neural Networks (Computer)
Multilayer neural networks
Databases
Sensitivity and Specificity

Cite this

@inproceedings{9eb1ae81dff54879acc2fc93dfe4eb14,
title = "Classifying ischemic events using a bayesian inference multilayer perceptron and input variable evaluation using automatic relevance determination",
abstract = "In this paper we present a Bayesian inference Multilayer Perceptron (MLP) which was used to classify the events of the Long Term ST Database (LTSTDB) as ischaemic or non-ischaemic episodes with an accuracy of 89.1{\%}, sensitivity of 82.3{\%} and specificity of 91.2{\%} when the accuracy of the winning paper was 90.7{\%}. The Automatic Relevance Determination (ARD) method was used to identify which of the extracted features that were used as input in the Bayesian inference MLP were the most important with respect to the models performance. ARD indicated that AT, a combination of the ST deviation and the duration of the episode, inspired from Langley et al [1], was the most important feature for determining Ischaemic episodes, given the data. A simple MLP which had as input variable of only AT was trained to verify the results of the ARD method. The classification accuracy was 85.8{\%} on the test set. We can conclude from the results that the most important extracted feature was ΔT.",
author = "Smyrnakis, {M. G.} and Evans, {D. J.}",
year = "2007",
month = "12",
day = "1",
doi = "10.1109/CIC.2007.4745482",
language = "English",
isbn = "9781424425334",
volume = "34",
pages = "305--308",
booktitle = "Computers in Cardiology",

}

Smyrnakis, MG & Evans, DJ 2007, Classifying ischemic events using a bayesian inference multilayer perceptron and input variable evaluation using automatic relevance determination. in Computers in Cardiology. vol. 34, 4745482, pp. 305-308, Computers in Cardiology 2007, CAR 2007, Durham, NC, United Kingdom, 30/09/07. https://doi.org/10.1109/CIC.2007.4745482

Classifying ischemic events using a bayesian inference multilayer perceptron and input variable evaluation using automatic relevance determination. / Smyrnakis, M. G.; Evans, D. J.

Computers in Cardiology. Vol. 34 2007. p. 305-308 4745482.

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

TY - GEN

T1 - Classifying ischemic events using a bayesian inference multilayer perceptron and input variable evaluation using automatic relevance determination

AU - Smyrnakis, M. G.

AU - Evans, D. J.

PY - 2007/12/1

Y1 - 2007/12/1

N2 - In this paper we present a Bayesian inference Multilayer Perceptron (MLP) which was used to classify the events of the Long Term ST Database (LTSTDB) as ischaemic or non-ischaemic episodes with an accuracy of 89.1%, sensitivity of 82.3% and specificity of 91.2% when the accuracy of the winning paper was 90.7%. The Automatic Relevance Determination (ARD) method was used to identify which of the extracted features that were used as input in the Bayesian inference MLP were the most important with respect to the models performance. ARD indicated that AT, a combination of the ST deviation and the duration of the episode, inspired from Langley et al [1], was the most important feature for determining Ischaemic episodes, given the data. A simple MLP which had as input variable of only AT was trained to verify the results of the ARD method. The classification accuracy was 85.8% on the test set. We can conclude from the results that the most important extracted feature was ΔT.

AB - In this paper we present a Bayesian inference Multilayer Perceptron (MLP) which was used to classify the events of the Long Term ST Database (LTSTDB) as ischaemic or non-ischaemic episodes with an accuracy of 89.1%, sensitivity of 82.3% and specificity of 91.2% when the accuracy of the winning paper was 90.7%. The Automatic Relevance Determination (ARD) method was used to identify which of the extracted features that were used as input in the Bayesian inference MLP were the most important with respect to the models performance. ARD indicated that AT, a combination of the ST deviation and the duration of the episode, inspired from Langley et al [1], was the most important feature for determining Ischaemic episodes, given the data. A simple MLP which had as input variable of only AT was trained to verify the results of the ARD method. The classification accuracy was 85.8% on the test set. We can conclude from the results that the most important extracted feature was ΔT.

UR - http://www.scopus.com/inward/record.url?scp=62949145742&partnerID=8YFLogxK

U2 - 10.1109/CIC.2007.4745482

DO - 10.1109/CIC.2007.4745482

M3 - Conference contribution

AN - SCOPUS:62949145742

SN - 9781424425334

VL - 34

SP - 305

EP - 308

BT - Computers in Cardiology

ER -