A comparative study of data science techniques based on ensemble classification algorithms in healthcare systems (Case study: Diabetic patients)
Subject Areas : Stochastic Models and SimulationFarnoosh Bagheri 1 , Reza Kamran Rad 2 , Morteza Darvishi 3
1 - Department of quantitative studies, university of canada west, Canada
2 - Industrial engineering Department, Faculty of Engineering, University of Semnan, Semnan, Iran, PostalCode35131-19111
3 - Department of industrial engineering, faculty of engineering, semnan university, semnan, iran
Keywords: Data mining, mixed classification techniques, healthcare system, Diabetes,
Abstract :
The adoption of unhealthy lifestyles by individuals can lead to the development of various health conditions, including hypertension, high blood fat levels, and diabetes, posing significant risks to their well-being. This study focuses on examining the lifestyle of patients with diabetes and high blood fat levels in the city of "bordekhoon," conducted at Health Care Centers. Diabetes is a global health concern that is rapidly increasing and is associated with substantial costs. By applying data mining techniques, early detection of diabetes can be achieved, which can help prevent the progression of the disease and mitigate complications such as cardiovascular issues, vision problems, and kidney diseases. Nowadays, data mining-based approaches are employed to predict diabetes and hypertension, aiming to enhance early diagnosis accuracy and obtain valuable insights. In this paper, a combination of classification techniques (Ensemble Method) is used to predict and identify two types of diabetes. Factors such as gender, diet, fasting plasma glucose (FPG), physical activity, cigarette consumption, age, genetic predisposition, and body mass index (BMI) are modeled and analyzed using IBM SPSS Modeler 18 software. The accuracy of the employed techniques is ultimately presented.