Probabilistic Grammar-based Neuroevolution for Physiological Signal Classification of Ventricular Tachycardia

Pak-kan WONG, Kwong-sak LEUNG, Man-leung WONG

Research output: Journal PublicationsJournal Article (refereed)Researchpeer-review

Abstract

Ventricular tachycardia is a rapid heart rhythm that begins in the lower chambers of the heart. When it happens continuously, this may result in life-threatening cardiac arrest. In this paper, we apply deep learning techniques to tackle the problem of the physiological signal classification of ventricular tachycardia, since deep learning techniques can attain outstanding performance in many medical applications. Nevertheless, human engineers are required to manually design deep neural networks to handle different tasks. This can be challenging because of many possible deep neural network structures. Therefore, a method, called ADAG-DNE, is presented to automatically design deep neural network structures using deep neuroevolution. Our approach defines a set of structures using probabilistic grammar and searches for best network structures using Probabilistic Model Building Genetic Programming. ADAG-DNE takes advantages of the probabilistic dependencies found among the structures of networks. When applying ADAG-DNE to the classification problem, our discovered model achieves better accuracy than AlexNet, ResNet, and seven non-neural network classifiers. It also uses about 2% of parameters of AlexNet, which means the inference can be made quickly. To summarize, our method evolves a deep neural network, which can be implemented in expert systems. The deep neural network achieves high accuracy. Moreover, it is simpler than existing deep neural networks. Thus, computational efficiency and diagnosis accuracy of the expert system can be improved.
Original languageEnglish
Pages (from-to)237-248
JournalExpert Systems with Applications
Volume135
Early online date6 Jun 2019
DOIs
Publication statusE-pub ahead of print - 6 Jun 2019

Fingerprint

Expert systems
Genetic programming
Medical applications
Computational efficiency
Deep neural networks
Classifiers
Engineers
Deep learning
Statistical Models

Bibliographical note

This research is supported by the Lingnan University Direct Grant DR16A7.

Keywords

  • Physiological signal classification
  • Heart diseaseNeuroevolution
  • Probabilistic grammar
  • Genetic programming
  • Deep neural network
  • Neuroevolution
  • Heart disease

Cite this

@article{015211566b66458cbae44d303731b9c9,
title = "Probabilistic Grammar-based Neuroevolution for Physiological Signal Classification of Ventricular Tachycardia",
abstract = "Ventricular tachycardia is a rapid heart rhythm that begins in the lower chambers of the heart. When it happens continuously, this may result in life-threatening cardiac arrest. In this paper, we apply deep learning techniques to tackle the problem of the physiological signal classification of ventricular tachycardia, since deep learning techniques can attain outstanding performance in many medical applications. Nevertheless, human engineers are required to manually design deep neural networks to handle different tasks. This can be challenging because of many possible deep neural network structures. Therefore, a method, called ADAG-DNE, is presented to automatically design deep neural network structures using deep neuroevolution. Our approach defines a set of structures using probabilistic grammar and searches for best network structures using Probabilistic Model Building Genetic Programming. ADAG-DNE takes advantages of the probabilistic dependencies found among the structures of networks. When applying ADAG-DNE to the classification problem, our discovered model achieves better accuracy than AlexNet, ResNet, and seven non-neural network classifiers. It also uses about 2{\%} of parameters of AlexNet, which means the inference can be made quickly. To summarize, our method evolves a deep neural network, which can be implemented in expert systems. The deep neural network achieves high accuracy. Moreover, it is simpler than existing deep neural networks. Thus, computational efficiency and diagnosis accuracy of the expert system can be improved.",
keywords = "Physiological signal classification, Heart diseaseNeuroevolution, Probabilistic grammar, Genetic programming, Deep neural network, Neuroevolution, Heart disease",
author = "Pak-kan WONG and Kwong-sak LEUNG and Man-leung WONG",
note = "This research is supported by the Lingnan University Direct Grant DR16A7.",
year = "2019",
month = "6",
day = "6",
doi = "10.1016/j.eswa.2019.06.012",
language = "English",
volume = "135",
pages = "237--248",
journal = "Expert Systems with Applications",
issn = "0957-4174",
publisher = "Elsevier Ltd",

}

Probabilistic Grammar-based Neuroevolution for Physiological Signal Classification of Ventricular Tachycardia. / WONG, Pak-kan; LEUNG, Kwong-sak; WONG, Man-leung.

In: Expert Systems with Applications, Vol. 135, 11.2019, p. 237-248.

Research output: Journal PublicationsJournal Article (refereed)Researchpeer-review

TY - JOUR

T1 - Probabilistic Grammar-based Neuroevolution for Physiological Signal Classification of Ventricular Tachycardia

AU - WONG, Pak-kan

AU - LEUNG, Kwong-sak

AU - WONG, Man-leung

N1 - This research is supported by the Lingnan University Direct Grant DR16A7.

PY - 2019/6/6

Y1 - 2019/6/6

N2 - Ventricular tachycardia is a rapid heart rhythm that begins in the lower chambers of the heart. When it happens continuously, this may result in life-threatening cardiac arrest. In this paper, we apply deep learning techniques to tackle the problem of the physiological signal classification of ventricular tachycardia, since deep learning techniques can attain outstanding performance in many medical applications. Nevertheless, human engineers are required to manually design deep neural networks to handle different tasks. This can be challenging because of many possible deep neural network structures. Therefore, a method, called ADAG-DNE, is presented to automatically design deep neural network structures using deep neuroevolution. Our approach defines a set of structures using probabilistic grammar and searches for best network structures using Probabilistic Model Building Genetic Programming. ADAG-DNE takes advantages of the probabilistic dependencies found among the structures of networks. When applying ADAG-DNE to the classification problem, our discovered model achieves better accuracy than AlexNet, ResNet, and seven non-neural network classifiers. It also uses about 2% of parameters of AlexNet, which means the inference can be made quickly. To summarize, our method evolves a deep neural network, which can be implemented in expert systems. The deep neural network achieves high accuracy. Moreover, it is simpler than existing deep neural networks. Thus, computational efficiency and diagnosis accuracy of the expert system can be improved.

AB - Ventricular tachycardia is a rapid heart rhythm that begins in the lower chambers of the heart. When it happens continuously, this may result in life-threatening cardiac arrest. In this paper, we apply deep learning techniques to tackle the problem of the physiological signal classification of ventricular tachycardia, since deep learning techniques can attain outstanding performance in many medical applications. Nevertheless, human engineers are required to manually design deep neural networks to handle different tasks. This can be challenging because of many possible deep neural network structures. Therefore, a method, called ADAG-DNE, is presented to automatically design deep neural network structures using deep neuroevolution. Our approach defines a set of structures using probabilistic grammar and searches for best network structures using Probabilistic Model Building Genetic Programming. ADAG-DNE takes advantages of the probabilistic dependencies found among the structures of networks. When applying ADAG-DNE to the classification problem, our discovered model achieves better accuracy than AlexNet, ResNet, and seven non-neural network classifiers. It also uses about 2% of parameters of AlexNet, which means the inference can be made quickly. To summarize, our method evolves a deep neural network, which can be implemented in expert systems. The deep neural network achieves high accuracy. Moreover, it is simpler than existing deep neural networks. Thus, computational efficiency and diagnosis accuracy of the expert system can be improved.

KW - Physiological signal classification

KW - Heart diseaseNeuroevolution

KW - Probabilistic grammar

KW - Genetic programming

KW - Deep neural network

KW - Neuroevolution

KW - Heart disease

UR - http://www.scopus.com/inward/record.url?scp=85067209517&partnerID=8YFLogxK

U2 - 10.1016/j.eswa.2019.06.012

DO - 10.1016/j.eswa.2019.06.012

M3 - Journal Article (refereed)

VL - 135

SP - 237

EP - 248

JO - Expert Systems with Applications

JF - Expert Systems with Applications

SN - 0957-4174

ER -