Diabetes risk prediction system and data visualization / Azizah Mohamad Imran and Hawa Mohd Ekhsan

Diabetes is a deadly chronic disease that has a negative impact on the entire body system. This disease affects millions of people, and a significant number of patients die because of its side effects each year. Undiagnosed diabetes can lead to nerve and kidney damage, heart and blood vessel disease, slow wound healing, hearing loss, and a variety of skin diseases. Moreover, the rapid growth of diabetes is very alarming and the need to identify the significant factor that leads to diabetes is increasing. Therefore, an efficient way to predict diabetics is required so that necessary procedures can be implemented ahead of time. A diabetes prediction system is implemented for predicting diabetes and visualizing the significant factors that lead to diabetes. The target users for this system are medical practitioners, individuals working in diabetes research centers, and the government. Secondary data has been used for this research. HTML, CSS, Python, and data visualization techniques are used to design the system. The overall development process is divided into four phases: planning, analysis, development, and testing. To determine Diabetes, the prediction model used and compared different machine learning algorithms such as Logistic Regression (LR) and Support Vector Machine (SVM). As a result, Logistic Regression has been selected as the prediction model because it displays the highest accuracy score. According to the usability testing evaluation, many respondents were satisfied with the system's usability.