A Decision Support System for Diabetes Prediction Using Machine Learning and Deep Learning Techniques

Yahyaoui A., Jamil A., Rasheed J., YEŞİLTEPE M.

1st International Informatics and Software Engineering Conference, IISEC 2019, Ankara, Turkey, 6 - 07 November 2019 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/ubmyk48245.2019.8965556
  • City: Ankara
  • Country: Turkey
  • Yıldız Technical University Affiliated: Yes


© 2019 IEEE.With the continuing increase in the number of the deadly diseases that threaten both human health and life, medical Decision Support Systems (DSS) continue to prove their effectiveness in providing physicians and other healthcare professionals with support in clinical decision making. Among these dangerous diseases, diabetes continues to be one of the leading one that has caused several deaths in the world. It is characterized by an increase in blood sugar levels which can have severe effects on other human organs. According to the International Diabetes Federation (IDA), 382 million people are living with diabetes and by 2035, these statistics will double to reach 592 million. In this paper, we propose a DSS for diabetes prediction based on Machine Learning (ML) techniques. We compared conventional machine learning with deep learning approaches. For conventional machine learning method, we considered the most commonly used classifiers: Support Vector Machine (SVM) and the Random Forest(RF). On the other hand, for Deep Learning (DL) we employed a fully Convolutional Neural Network (CNN) to predict and detect the diabetes patients. The proposed system is evaluated on publicly available Pima Indians Diabetes database which consisted of total 768 samples each with 8 features. 500 samples were labeled as non-diabetic while 268 were diabetic patients. The overall accuracy obtained using DL, SVM and RF was 76.81%, 65.38% and 83.67% respectively. The experimental results show that RF was more effective for diabetes prediction compared to deep learning and SVM methods.