Portal de Periódicos da CAPES

Study and Impact Analysis of Machine Learning Approaches for Smart Healthcare in Predicting Mellitus Diabetes on Clinical Data

2021; Springer Nature; Linguagem: Inglês

10.1007/978-981-16-5304-9_7

1868-4408

Aisworya Mohanty, Sasmita Parida, Suvendu Chandan Nayak, Bibudhendu Pati, Chhabi Rani Panigrahi,

COVID-19 diagnosis using AI

Diabetes is a chronic disease that results in too much glucose in the blood. In current scenario it becomes a communal health crisis in human beings. Sometimes it causes saviour injuries to mammals leading them to death. To evade loss of precious lives several machine learning methodologies are implemented for the prediction of diabetes. The task of predictive analytics is very challenging but it can eventually assist the physicians to predict the disease and notify them about the victim’s condition. The victim could be cured by early treatment or by taking precautions before the patient is being attacked. In this research work six different types of supervised machine learning (ML) algorithms are used to solve this hazardous dilemma such as “Support Vector Machine (SVM)”, “Decision Tree (DT)”, “Artificial Neural Network (ANN)”, “Random Forest (RF)”, “K-Nearest Neighbor (KNN)” and “Logistic Regression (LR)” are applied on the “Pima Indians (PIDD) dataset”. On the basis of accuracy precision the effectiveness of the classifier is evaluated. For experimental purpose, the details of a patient’s medical record are applied on the dataset. By discerning this medical record a scholar can inspect the symptoms and can effortlessly predict the stage to which the patient belongs. By this early extrapolation the patient can acquire some recommended medication related to the life risky syndrome to ensure the safety of the life of the patient shortly. Smart healthcare includes smartly handle the patient from past data analysis for predicting chronic diabetes disease.