Machine learning technique for detecting diabetes disease

Abstract

Diabetes is considered a critical disease and it has been a growing concern owing to its increased morbidity. Moreover, the average age of people who are affected by diabetes illness has currently declined to the mid-20s. Given the high prevalence, it is necessary to address this problem effectively. Due to the significant prevalence of diabetes illness, it is essential to handle and address this issue appropriately. Currently, machine learning methods are considered a vital area for detecting and diagnosing disease. These methods can learn from data and classify data based on the coordinate subjects. This paper presents a model for detecting diabetes illness based on a machine learning technique. The Support Vector Machine (SVM) algorithm is used for classifying the people who are categorized as patients with diabetes disease from the people who are categorized as non-diabetic. Further, the database is compiled from the Pima Indian Diabetes Dataset (PIDD). The results show that the proposed model achieves 81.8% accuracy. Moreover, the proposed model achieves 84.34% sensitivity and 74.35% specificity.