Quantum hybrid technique for feature selection in classification of cervical cells

Abdullah M. Iliyasu; Chastine Fatichah; Ahmed S. Salama; Jinhua She

Quantum hybrid technique for feature selection in classification of cervical cells

Abdullah M. Iliyasu^*
, Chastine Fatichah
, Ahmed S. Salama
, Jinhua She

^*Corresponding author for this work

Prince Sattam Bin Abdulaziz University
Tokyo University of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Citation (Scopus)

Abstract

A quantum hybrid technique (QHT) that blends the adaptive search capability of the quantum-behaved particle swarm optimisation (QPSO) method with the intuitionistic rationality inherent to traditional fuzzy k-nearest neighbours (Fuzzy k-NN) algorithm is proposed for efficient feature selection and classification of cells in cervical smeared (CS) images. Using a dataset of almost 1000 images, the proposed QHT technique drastically reduces the number of cell features, which leads to an average cell classification accuracy of 90%. This is an average of 7% improvement in the accuracy obtainable without the QPSO feature selection of the proposed QHT technique. With additional modifications, the proposed technique could prove useful in cervical cancer detection and diagnosis.

Original language	English
Title of host publication	ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications
Publisher	Fuji Technology Press
ISBN (Electronic)	9784990534349
Publication status	Published - 2016
Event	7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016 - Beijing, China Duration: 3 Nov 2016 → 6 Nov 2016

Publication series

Name	ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications

Conference

Conference	7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016
Country/Territory	China
City	Beijing
Period	3/11/16 → 6/11/16

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Cervical cancer
Computational intelligence
Fuzzy K-NN
Hybrid intelligent techniques
Medical image processing
QPSO

Cite this

Iliyasu, A. M., Fatichah, C., Salama, A. S., & She, J. (2016). Quantum hybrid technique for feature selection in classification of cervical cells. In ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications (ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications). Fuji Technology Press.

Iliyasu, Abdullah M. ; Fatichah, Chastine ; Salama, Ahmed S. et al. / Quantum hybrid technique for feature selection in classification of cervical cells. ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications. Fuji Technology Press, 2016. (ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications).

@inproceedings{e801884d0715438f9561b1fc90df9848,

title = "Quantum hybrid technique for feature selection in classification of cervical cells",

abstract = "A quantum hybrid technique (QHT) that blends the adaptive search capability of the quantum-behaved particle swarm optimisation (QPSO) method with the intuitionistic rationality inherent to traditional fuzzy k-nearest neighbours (Fuzzy k-NN) algorithm is proposed for efficient feature selection and classification of cells in cervical smeared (CS) images. Using a dataset of almost 1000 images, the proposed QHT technique drastically reduces the number of cell features, which leads to an average cell classification accuracy of 90\%. This is an average of 7\% improvement in the accuracy obtainable without the QPSO feature selection of the proposed QHT technique. With additional modifications, the proposed technique could prove useful in cervical cancer detection and diagnosis.",

keywords = "Cervical cancer, Computational intelligence, Fuzzy K-NN, Hybrid intelligent techniques, Medical image processing, QPSO",

author = "Iliyasu, \{Abdullah M.\} and Chastine Fatichah and Salama, \{Ahmed S.\} and Jinhua She",

note = "Publisher Copyright: {\textcopyright} 2016, Fuji Technology Press. All rights reserved.; 7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016 ; Conference date: 03-11-2016 Through 06-11-2016",

year = "2016",

language = "English",

series = "ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications",

publisher = "Fuji Technology Press",

booktitle = "ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications",

address = "Japan",

}

Iliyasu, AM, Fatichah, C, Salama, AS & She, J 2016, Quantum hybrid technique for feature selection in classification of cervical cells. in ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications. ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications, Fuji Technology Press, 7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016, Beijing, China, 3/11/16.

Quantum hybrid technique for feature selection in classification of cervical cells. / Iliyasu, Abdullah M.; Fatichah, Chastine; Salama, Ahmed S. et al.
ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications. Fuji Technology Press, 2016. (ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Quantum hybrid technique for feature selection in classification of cervical cells

AU - Iliyasu, Abdullah M.

AU - Fatichah, Chastine

AU - Salama, Ahmed S.

AU - She, Jinhua

PY - 2016

Y1 - 2016

N2 - A quantum hybrid technique (QHT) that blends the adaptive search capability of the quantum-behaved particle swarm optimisation (QPSO) method with the intuitionistic rationality inherent to traditional fuzzy k-nearest neighbours (Fuzzy k-NN) algorithm is proposed for efficient feature selection and classification of cells in cervical smeared (CS) images. Using a dataset of almost 1000 images, the proposed QHT technique drastically reduces the number of cell features, which leads to an average cell classification accuracy of 90%. This is an average of 7% improvement in the accuracy obtainable without the QPSO feature selection of the proposed QHT technique. With additional modifications, the proposed technique could prove useful in cervical cancer detection and diagnosis.

AB - A quantum hybrid technique (QHT) that blends the adaptive search capability of the quantum-behaved particle swarm optimisation (QPSO) method with the intuitionistic rationality inherent to traditional fuzzy k-nearest neighbours (Fuzzy k-NN) algorithm is proposed for efficient feature selection and classification of cells in cervical smeared (CS) images. Using a dataset of almost 1000 images, the proposed QHT technique drastically reduces the number of cell features, which leads to an average cell classification accuracy of 90%. This is an average of 7% improvement in the accuracy obtainable without the QPSO feature selection of the proposed QHT technique. With additional modifications, the proposed technique could prove useful in cervical cancer detection and diagnosis.

KW - Cervical cancer

KW - Computational intelligence

KW - Fuzzy K-NN

KW - Hybrid intelligent techniques

KW - Medical image processing

KW - QPSO

UR - https://www.scopus.com/pages/publications/84997794594

M3 - Conference contribution

AN - SCOPUS:84997794594

T3 - ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications

BT - ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications

PB - Fuji Technology Press

T2 - 7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016

Y2 - 3 November 2016 through 6 November 2016

ER -

Quantum hybrid technique for feature selection in classification of cervical cells

Abstract

Publication series

Conference

UN SDGs

Keywords

Other files and links

Fingerprint

Cite this