Heterogeneous ensemble pruning based on Bee Algorithm for mammogram classification

Ashwaq Qasem, Shahnorbanun Sahran, Siti Norul Huda Sheikh Abdullah, Dheeb Albashish, Rizuana Iqbal Hussain, Shantini Arasaratnam

Research output: Contribution to journalArticle

Abstract

In mammogram, masses are primary indication of breast cancer; and it is necessary to classify them as malignant or benign. In this classification task, Computer Aided Diagnostic (CAD) system by using ensemble learning is able to assist radiologists to have better diagnosis of mammogram images. Ensemble learning consists of two steps, generating multiple base classifiers and then combining them together. However, combining all base classifier in the ensemble model increases the computational cost and time. Therefore, ensemble pruning is an important step in ensemble learning to select the ensemble's members. Due to huge subsets of combination in the ensemble, selecting the proper ensemble subset is desirable. The objective of this paper is to select the optimal ensemble subset by using Bee Algorithm (BA). A pool of different classifier models such as Support vector machine, k-nearest neighbour and linear discriminant analysis classifiers have been trained using different samples of training data and 12 groups of features. Then, by using this pool of classifier models, BA was used to exploit and explore random uniform combination subsets of the trained classifiers. As a result, the best subset will be selected as the optimal ensemble. The mammogram image dataset that was used to test the model has been collected from Hospital Kuala Lumpur (HKL) and consists of 263 benign and malignant masses. The proposed method gives 77 % of Area Under Curve(AUC), 83% of accuracy, 93% of specificity and 60% of sensitivity.

Original languageEnglish
Pages (from-to)231-239
Number of pages9
JournalInternational Journal of Advanced Computer Science and Applications
Volume9
Issue number12
DOIs
Publication statusPublished - 1 Jan 2018

Fingerprint

Classifiers
Discriminant analysis
Set theory
Support vector machines
Costs

Keywords

  • Bee algorithm
  • Breast cancer
  • Ensemble learning
  • Ensemble pruning
  • Mammogram

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Heterogeneous ensemble pruning based on Bee Algorithm for mammogram classification. / Qasem, Ashwaq; Sahran, Shahnorbanun; Sheikh Abdullah, Siti Norul Huda; Albashish, Dheeb; Iqbal Hussain, Rizuana; Arasaratnam, Shantini.

In: International Journal of Advanced Computer Science and Applications, Vol. 9, No. 12, 01.01.2018, p. 231-239.

Research output: Contribution to journalArticle

@article{828153503c044765919307ae69779853,
title = "Heterogeneous ensemble pruning based on Bee Algorithm for mammogram classification",
abstract = "In mammogram, masses are primary indication of breast cancer; and it is necessary to classify them as malignant or benign. In this classification task, Computer Aided Diagnostic (CAD) system by using ensemble learning is able to assist radiologists to have better diagnosis of mammogram images. Ensemble learning consists of two steps, generating multiple base classifiers and then combining them together. However, combining all base classifier in the ensemble model increases the computational cost and time. Therefore, ensemble pruning is an important step in ensemble learning to select the ensemble's members. Due to huge subsets of combination in the ensemble, selecting the proper ensemble subset is desirable. The objective of this paper is to select the optimal ensemble subset by using Bee Algorithm (BA). A pool of different classifier models such as Support vector machine, k-nearest neighbour and linear discriminant analysis classifiers have been trained using different samples of training data and 12 groups of features. Then, by using this pool of classifier models, BA was used to exploit and explore random uniform combination subsets of the trained classifiers. As a result, the best subset will be selected as the optimal ensemble. The mammogram image dataset that was used to test the model has been collected from Hospital Kuala Lumpur (HKL) and consists of 263 benign and malignant masses. The proposed method gives 77 {\%} of Area Under Curve(AUC), 83{\%} of accuracy, 93{\%} of specificity and 60{\%} of sensitivity.",
keywords = "Bee algorithm, Breast cancer, Ensemble learning, Ensemble pruning, Mammogram",
author = "Ashwaq Qasem and Shahnorbanun Sahran and {Sheikh Abdullah}, {Siti Norul Huda} and Dheeb Albashish and {Iqbal Hussain}, Rizuana and Shantini Arasaratnam",
year = "2018",
month = "1",
day = "1",
doi = "10.14569/IJACSA.2018.091234",
language = "English",
volume = "9",
pages = "231--239",
journal = "International Journal of Advanced Computer Science and Applications",
issn = "2158-107X",
publisher = "Science and Information Organization",
number = "12",

}

TY - JOUR

T1 - Heterogeneous ensemble pruning based on Bee Algorithm for mammogram classification

AU - Qasem, Ashwaq

AU - Sahran, Shahnorbanun

AU - Sheikh Abdullah, Siti Norul Huda

AU - Albashish, Dheeb

AU - Iqbal Hussain, Rizuana

AU - Arasaratnam, Shantini

PY - 2018/1/1

Y1 - 2018/1/1

N2 - In mammogram, masses are primary indication of breast cancer; and it is necessary to classify them as malignant or benign. In this classification task, Computer Aided Diagnostic (CAD) system by using ensemble learning is able to assist radiologists to have better diagnosis of mammogram images. Ensemble learning consists of two steps, generating multiple base classifiers and then combining them together. However, combining all base classifier in the ensemble model increases the computational cost and time. Therefore, ensemble pruning is an important step in ensemble learning to select the ensemble's members. Due to huge subsets of combination in the ensemble, selecting the proper ensemble subset is desirable. The objective of this paper is to select the optimal ensemble subset by using Bee Algorithm (BA). A pool of different classifier models such as Support vector machine, k-nearest neighbour and linear discriminant analysis classifiers have been trained using different samples of training data and 12 groups of features. Then, by using this pool of classifier models, BA was used to exploit and explore random uniform combination subsets of the trained classifiers. As a result, the best subset will be selected as the optimal ensemble. The mammogram image dataset that was used to test the model has been collected from Hospital Kuala Lumpur (HKL) and consists of 263 benign and malignant masses. The proposed method gives 77 % of Area Under Curve(AUC), 83% of accuracy, 93% of specificity and 60% of sensitivity.

AB - In mammogram, masses are primary indication of breast cancer; and it is necessary to classify them as malignant or benign. In this classification task, Computer Aided Diagnostic (CAD) system by using ensemble learning is able to assist radiologists to have better diagnosis of mammogram images. Ensemble learning consists of two steps, generating multiple base classifiers and then combining them together. However, combining all base classifier in the ensemble model increases the computational cost and time. Therefore, ensemble pruning is an important step in ensemble learning to select the ensemble's members. Due to huge subsets of combination in the ensemble, selecting the proper ensemble subset is desirable. The objective of this paper is to select the optimal ensemble subset by using Bee Algorithm (BA). A pool of different classifier models such as Support vector machine, k-nearest neighbour and linear discriminant analysis classifiers have been trained using different samples of training data and 12 groups of features. Then, by using this pool of classifier models, BA was used to exploit and explore random uniform combination subsets of the trained classifiers. As a result, the best subset will be selected as the optimal ensemble. The mammogram image dataset that was used to test the model has been collected from Hospital Kuala Lumpur (HKL) and consists of 263 benign and malignant masses. The proposed method gives 77 % of Area Under Curve(AUC), 83% of accuracy, 93% of specificity and 60% of sensitivity.

KW - Bee algorithm

KW - Breast cancer

KW - Ensemble learning

KW - Ensemble pruning

KW - Mammogram

UR - http://www.scopus.com/inward/record.url?scp=85059509166&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059509166&partnerID=8YFLogxK

U2 - 10.14569/IJACSA.2018.091234

DO - 10.14569/IJACSA.2018.091234

M3 - Article

VL - 9

SP - 231

EP - 239

JO - International Journal of Advanced Computer Science and Applications

JF - International Journal of Advanced Computer Science and Applications

SN - 2158-107X

IS - 12

ER -