Abstract:
Glucose, commonly called sugar, is an important source of energy needed by all the cells and organs of our body. High blood sugar or glucose levels continue to lead to diabetes, In the field of human health, computer vision plays a major role in reducing human judgment and providing accurate results. The main purpose of this study is to provide better stages of diabetes. This manuscript focuses on studying comparative analysis of algorithms to improve the accuracy of the prediction model using machine learning and data mining techniques. Pima Indians Diabetes Dataset standard based on the UCI machine learning curve has been used. Selections are made to maximize the power of the data set. Algorithms such as Support Vector Machine (SVM), Naive Bayes, Decision Tree, K-Nearest Neighbors (KNN) and Logistic Regression are used in the database to test model performance. Test metrics like Precision, Remember, Specification and total error for each model are calculated to suggest the best sample data separator. Red to the location for information The analysis tool kit was used to compare 78% accuracy in the SVM algorithm. 2 (T2D).
