Diabetes Mellitus Classification Using Hybrid Machine Learning With Stacking Technique

Abdulrazak Yahya, Saleh and Bobby Brixtone, Batou (2022) Diabetes Mellitus Classification Using Hybrid Machine Learning With Stacking Technique. In: The 2nd International Conference On Emerging Smart Technologies And Applications (eSmarTA2022), 25 -26 October 2022, Virtual, Online.

PDF Diabetes Mellitus - Copy.pdf Download (141kB)

Abstract

Diabetes Mellitus is one of the world's fastest growing and most fatal diseases. Diabetes mellitus is a metabolic disorder that worsens as the body's ability to metabolise glucose declines. Machine learning classifiers can aid in disease prediction or detection based on the severity of the patient's symptoms. This work proposed a new model, a hybrid machine learning model with stacking classifier technique. The study makes use of a Chinese diabetes dataset that includes over 100,000 people of diverse races and other features. The data was pre-processed by using means to replace and eliminate missing values, and the model's performance will be improved by using the stacking classifier technique. The study found that the proposed model performed better in classifying diabetic mellitus illnesses, with results of 0.914 percent, 0.926 percent, 0.914 percent and 0.914 percent in accuracy, precision, recall, and F1 score, respectively. When mixing several types of models, the stacking classifier technique produced better results.

Actions (For repository members only: login required)

View Item