Improving Neural Network Performance with Feature Selection Using Pearson Correlation Method for Diabetes Disease Detection

April Firman Daru; Mohammad Burhan Hanif; Edi Widodo

doi:10.30595/juita.v9i1.9941

Authors

April Firman Daru Universitas Semarang
Mohammad Burhan Hanif Universitas Semarang
Edi Widodo

DOI:

https://doi.org/10.30595/juita.v9i1.9941

Keywords:

Neural Network, Pearson Correlation, Diabetes.

Abstract

Diabetic or silent killer diseases are an alarming scourge for the world and are classed as serious diseases. In Indonesia, the increase in diabetics occurred by 2% in vulnerable times between 2013 to 2018. This affects all sectors, both medical services and the financial sector. The Neural Network method as a data mining algorithm is present to overcome the burden that arises as an early detection analysis of the onset of disease. However, Neural Network has slow training capabilities and can identify important attributes in the data resulting in a decrease in performance. Pearson correlation is good at handling data with mixed-type attributes and is good at measuring information between attributes and attributes with labels. With this, the purpose of this study will be to use the Pearson correlation method as a selection of features to improve neural network performance in diabetes detection and measure the extent of accuracy obtained from the method. The dataset used is diabetes data 130-US hospital UCI with a record number of 101767 and the number of attributes as many as 50 attributes. The results of this study found that Pearson correlation can improve neural network accuracy performance from 94.93% to 96.00%. As for the evaluation results on the AUC value increased from 0.8077 to 0.8246. Thus Pearson's Correlation algorithm can work well for feature selection on neural network methods and can provide solutions to improved diabetes detection accuracy.

References

[1] N. Kumar and M. Abedin, “Comparative Approaches for Classification of Diabetes Mellitus Data: Machine Learning Paradigm,” Comput. Methods Programs Biomed., 2017.

[2] M. Chan, “Global Report On Diabetes,” World Heal. Organ., 2013.

[3] K. Kesehatan, “HARI DIABETES SEDUNIA TAHUN 2018.” Kesehatan, Infodatin pusat data dan informasi kementrian RI, 2019.

[4] M. Rigla, B. Pons, and M. Elena, “A web-based clinical decision support system for gestational diabetes : Automatic diet prescription and detection of insulin needs,” Int. J. Med. Inform., 2017.

[5] F. Mansourypoor and S. Asadi, “Development of a Reinforcement Learning-based Evolutionary Fuzzy Rule-Based System for Diabetes Diagnosis Fatemeh,” Comput. Biol. Med., 2017.

[6] N. Mehrbakhsh, I. Othman, D. Mohammad, A. Hossein, and S. Leila, “Accuracy Improvement for Diabetes Disease Classification A Case on a Public Medical Dataset,” Fuzzy Inf. Eng., 2017.

[7] S. Kang, “Personalized prediction of drug efficacy for diabetes treatment via patient-level sequential modeling with neural networks,” Artif. Intell. Med., pp. 2–7, 2018.

[8] C. J. Bennett, R. A. Stewart, and J. W. Lu, “Forecasting low voltage distribution network demand profiles using a pattern recognition based expert system,” Energy, vol. 67, pp. 200–212, 2014.

[9] G. C. Liao, “Hybrid improved differential evolution and wavelet neural network with load forecasting problem of air conditioning,” Int. J. Electr. Power Energy Syst., vol. 61, pp. 673–682, 2014.

[10] C. J. Lu, T. S. Lee, and C. M. Lian, “Sales forecasting for computer wholesalers: A comparison of multivariate adaptive regression splines and artificial neural networks,” Decis. Support Syst., vol. 54, no. 1, pp. 584–596, 2012.

[11] F. C. Cruz, E. F. S. Filho, M. C. S. Albuquerque, I. C. Silva, C. T. T. Farias, and L. L. Gouvêa, “Efficient feature selection for neural network based detection of flaws in steel welded joints using ultrasound testing,” Ultrasonics, vol. 73, pp. 1–8, 2017.

[12] Y. Mu, X. Liu, and L. Wang, “A Pearson’s correlation coefficient based decision tree and its parallel implementation,” Inf. Sci. (Ny)., vol. 435, pp. 40–58, 2018.

[13] Sugiyono, Metode Penelitian Kuantitatif, kualitatif dan R & D. Bandung: Alfabeta, 2012.

[14] B. Strack et al., “Diabetes Hospital,” 2020. https://archive.ics.uci.edu/ml/datasets/diabetes+130-us+hospitals+for+years+1999-2008 (accessed Sep. 23, 2020).

[15] H. Serge, “Analisis Regresi Dan Korelasi Antara Seleksi Bobot Badan Fase Starter Terhadap Produksi Ayam Ras Petelur Tipe Medium,” no. March, 2016.