A Homogeneous Ensemble Classifier for Breast Cancer Detection Using Parameters Tuning of MLP Neural Network

Figures & data

Table 1. Comparison of existing techniques for breast cancer detection

Reference	Methods	Advantages	Disadvantages	Results
Zhu et al. (2019)	Investigation of deep learning for differentiation between molecular subtypes of breast cancer	The C4.5 has been introduced as the most widely used approach to cancer diagnosis	Basic and statistical criteria were not used in the comparisons	-
Talatian Azad, Ahmadi, and Rezaeipanah (2021)	An intelligent ensemble classification method based on neural network	Stacked generalization-based ensemble classification improves model prediction results	Simultaneous optimization of parameters increases the complexity of the GA	98.74%
El Ouassif, Idri, and Hosni (2021)	Homogeneous ensemble based on four types of SVM classifiers	Using the Borda Count to rank the best classifiers	The effect of parameter tuning on classification performance has not been investigated	90.88%
Rezaeipanah and Ahmadi (2020)	Using multi-stage weight adjustment in the neural network	Find the best MLP structure by performing simultaneous and automatic search	There is no comparison with single classification models	98.74%
Wang et al. (2018)	An SVM-based ensemble algorithm	Using large real datasets to perspectives of model effectiveness	Feature selection techniques have not been used to reduce model complexity	97.89%
Kadam, Jadhav, and Vijayakumar (2019)	Using of feature ensemble learning based on stacked sparse autoencoders	Using the integration of different classifiers to benefit from the representation of all layers	Comparing the rules to make a final decision lacks statistical analysis	98.60%
Abdar et al. (2020)	A novel nested ensemble technique based on voting and stacking	Using the two-layer nesting ensemble technique improves the performance	No approach has been proposed to deal with high-dimensionality and large data volumes	98.07%
Idri, Hosni, and Abnane (2020)	Heterogeneous ensembles classification based on SVM, MLP, and ID3	Improving heterogeneous ensemble techniques developed based on the Borda Count method	The complexity is relatively high due to the lack of feature selection approaches	95.90%

Figure 1. The architecture of the proposed algorithm.

Figure 2. Flowchart of the proposed method.

Figure 3. Solutions representation structure.

Figure 4. Evaluation of different algorithms in tuning MLP parameters on the WBCD dataset.

Figure 5. Evaluation of different algorithms in tuning MLP parameters on the WDBC dataset.

Figure 6. Evaluation of different algorithms in tuning MLP parameters on the WPBC dataset.

Table 2. Performance of proposed algorithm with/without features selection

Criterion	With FS	Without FS	With FS	Without FS	With FS	Without FS
	WBCD		WDBC		WPBC
No. of selected features	4.8	9.0	15.5	31	12.1	34
No. of connections	36.7	100.4	83.4	308.5	68.0	325.8
Accuracy (%)	98.79	98.61	98.52	98.11	97.92	97.76
Sensitivity (%)	99.30	98.48	97.75	97.34	97.40	97.32
Specificity (%)	98.44	98.23	99.05	98.64	98.22	98.36
F1 score (%)	97.74	97.81	97.98	98.40	98.35	98.87
Runtime (s)	63.1	64.8	125.9	133.7	78.6	80.2

Table 3. Details of neural network configuration in mode with/without feature selection

Datasets	With features selection	Without features selection
WBCD	2 (2–3)	2 (3–3)
WDBC	2 (2–3)	3 (2 – 2 – 2)
WPBC	3 (2 – 2 – 1)	4 (2 – 2 – 1 – 2)

Table 4. Comparison of the proposed algorithm with other methods based on WBCD, WDBC, and WPBC datasets

Methods	WBCD dataset	Methods	WDBC dataset	Methods	WPBC dataset
SVM	96.87	SVM	94.85	SVM	79.08
MLP	97.00	MLP	95.11	MLP	81.13
NSGA-II (Ibrahim and Shamsuddin 2018)	97.02	WAUCE (Wang et al. 2018)	97.68	ANFIS+GA (Bilalović and Avdagić 2018)	90.01
RF+GA (Aličković and Subasi, 2017)	99.49	PSO-KDE (Ed-daoudy and Maalmi 2020)	98.45	PSO-KDE (Ed-daoudy and Maalmi 2020)	98.46
PSO-KDE (Sheikhpour, Sarram, and Sheikhpour 2016)	98.54	RF+KNN+SVM (Kumar and Poonkodi 2019)	95.61	GA+CFS+RF (Singla, Ghosh, and Kumari 2019)	86.37
SVM+AR (Ed-daoudy and Maalmi 2020)	98.01	PCA+LDA+ANNFIS (Preetha and Jinny 2020)	98.61	Xgboost (Narvekar et al. 2019)	97.72
Proposed algorithm	98.79	Proposed algorithm	98.52	Proposed algorithm	97.92

Table 5. Comparison of the proposed algorithm with other methods based on BCCD dataset

Basic classifiers	Accuracy (%)	Previous literature	Accuracy (%)
SVM (Li and Chen 2018)	71.40	FS-SVM (Singh 2019)	82.39
MLP (Li and Chen 2018)	60.00	Fuzzy classifier (Ontiveros-Robles and Melin 2020)	75.95
ID3 (Li and Chen 2018)	68.60	FS+k-NN (Abdel-Basset et al. 2020)	73.63
RF (Li and Chen 2018)	74.30	PCA+GRNN (Yavuz and Eyupoglu 2020)	97.73
Proposed algorithm	94.62	Proposed algorithm	94.62

Zhu, Z., E. Albadawy, A. Saha, J. Zhang, M. R. Harowicz, and M. A. Mazurowski. 2019. Deep learning for identifying radiogenomic associations in breast cancer. Computers in Biology and Medicine 109:85–90. doi:10.1016/j.compbiomed.2019.04.018.

 PubMed Web of Science ®Google Scholar

Talatian Azad, S., G. Ahmadi, and A. Rezaeipanah. 2021. An intelligent ensemble classification method based on multi-layer perceptron neural network and evolutionary algorithms for breast cancer diagnosis. Journal of Experimental & Theoretical Artificial Intelligence 1–21. In press. doi:10.1080/0952813X.2021.1938698.

 Web of Science ®Google Scholar

El Ouassif, B., A. Idri, and M. Hosni. 2021. Homogeneous ensemble based support vector machine in breast cancer diagnosis. In Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2021) , Vol. 5, pp. 352–60. Vienna, Austria.

 Google Scholar

Rezaeipanah, A., and G. Ahmadi. 2020. Breast cancer diagnosis using multi-stage weight adjustment in the MLP neural network. The Computer Journal. In press. doi:10.1093/comjnl/bxaa109.

 Web of Science ®Google Scholar

Wang, H., B. Zheng, S. W. Yoon, and H. S. Ko. 2018. A support vector machine-based ensemble algorithm for breast cancer diagnosis. European Journal of Operational Research 267 (2):687–99. doi:10.1016/j.ejor.2017.12.001.

 Web of Science ®Google Scholar

Kadam, V. J., S. M. Jadhav, and K. Vijayakumar. 2019. Breast cancer diagnosis using feature ensemble learning based on stacked sparse autoencoders and softmax regression. Journal of Medical Systems 43 (8):1–11. doi:10.1007/s10916-019-1397-z.

 Web of Science ®Google Scholar

Abdar, M., M. Zomorodi-Moghadam, X. Zhou, R. Gururajan, X. Tao, P. D. Barua, and R. Gururajan. 2020. A new nested ensemble technique for automated diagnosis of breast cancer. Pattern Recognition Letters 132:123–1875. doi:10.1016/j.patrec.2018.11.004.

 Web of Science ®Google Scholar

Idri, A., M. Hosni, and I. Abnane. 2020. Assessing the impact of parameters tuning in ensemble-based breast cancer classification. Health and Technology 10 (5):1239–55. doi:10.1007/s12553-020-00453-2.

 Web of Science ®Google Scholar

Ibrahim, A. O., and S. M. Shamsuddin. 2018. Intelligent breast cancer diagnosis based on enhanced Pareto optimal and multilayer perceptron neural network. International Journal of Computer Aided Engineering and Technology 10 (5):543–56. doi:10.1504/IJCAET.2018.094327.

 Google Scholar

Bilalović, O., and Z. Avdagić. 2018. Robust breast cancer classification based on GA optimized ANN and ANFIS-voting structures. In 2018 41st International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), 0279–84. IEEE, Opatija, Croatia, May.

 Google Scholar

Aličković, E., and A. Subasi. 2017. Breast cancer diagnosis using GA feature selection and Rotation Forest. Neural Computing and Applications 28 (4):753–63. doi:10.1007/s00521-015-2103-9.

 Web of Science ®Google Scholar

Ed-daoudy, A., and K. Maalmi. 2020. Breast cancer classification with reduced feature set using association rules and support vector machine. NetMAHIB 9 (1):34.

 Google Scholar

Sheikhpour, R., M. A. Sarram, and R. Sheikhpour. 2016. Particle swarm optimization for bandwidth determination and feature selection of kernel density estimation-based classifiers in diagnosis of breast cancer. Applied Soft Computing 40:113–31. doi:10.1016/j.asoc.2015.10.005.

 Web of Science ®Google Scholar

Kumar, A., and M. Poonkodi. 2019. Comparative study of different machine learning models for breast cancer diagnosis. In Innovations in soft computing and information technology, ed. Jayeeta Chattopadhyay, Rahul Singh and Vandana Bhattacherjee, 17–25. Singapore: Springer.

 Google Scholar

Singla, S., P. Ghosh, and U. Kumari. 2019. Breast cancer detection using genetic algorithm with correlation based feature selection: Experiment on different datasets. International Journal of Computer Sciences and Engineering 7 (4):406–10. doi:10.26438/ijcse/v7i4.406410.

 Google Scholar

Preetha, R., and S. V. Jinny. 2020. Early diagnose breast cancer with PCA-LDA based FER and neuro-fuzzy classification system. Journal of Ambient Intelligence and Humanized Computing 12:7195–204. doi:10.1007/s12652-020-02395-z.

 Web of Science ®Google Scholar

Narvekar, S. D., A. Patil, J. Patil, and S. Kudoo. 2019. Prognostication of breast cancer using data mining and machine learning. International Journal of Advance Research, Ideas and Innovations in Technology 5 (2):921–24.

 Google Scholar

Li, Y., and Z. Chen. 2018. Performance evaluation of machine learning methods for breast cancer prediction. Applied and Computational Mathematics 7 (4):212–16. doi:10.11648/j.acm.20180704.15.

 Google Scholar

Singh, B. K. 2019. Determining relevant biomarkers for prediction of breast cancer using anthropometric and clinical features: A comparative investigation in machine learning paradigm. Biocybernetics and Biomedical Engineering 39 (2):393–409. doi:10.1016/j.bbe.2019.03.001.

 Web of Science ®Google Scholar

Ontiveros-Robles, E., and P. Melin. 2020. Toward a development of general type-2 fuzzy classifiers applied in diagnosis problems through embedded type-1 fuzzy classifiers. Soft Computing 24 (1):83–99. doi:10.1007/s00500-019-04157-2.

 Web of Science ®Google Scholar

Abdel-Basset, M., D. El-Shahat, I. El-henawy, V. H. C. de Albuquerque, and S. Mirjalili. 2020. A new fusion of grey wolf optimizer algorithm with a two-phase mutation for feature selection. Expert Systems with Applications 139:112824. doi:10.1016/j.eswa.2019.112824.

 Web of Science ®Google Scholar

Yavuz, E., and C. Eyupoglu. 2020. An effective approach for breast cancer diagnosis based on routine blood analysis features. Medical & Biological Engineering & Computing 58:1583–601. doi:10.1007/s11517-020-02187-9.

 PubMed Web of Science ®Google Scholar