| Abstrakt: |
The primary purpose of this study is to use unique blood sample data to the analysis of the gender ratio for diabetes prediction. The Components and Procedures: Glucose level, diastolic bp, blood thickness, insulin levels, body mass index, age, and skin are some of the features included in the dataset utilised for training and testing the suggested prediction models. Patients' data were gathered from various online sources, and the framework was built using their novel blood sample data and the Voting Classifier over Support Vector Regression machine learning algorithm applied to a diabetes dataset that contains 1695 records with 10 attributes, the samples considered are (N=10) for both the algorithms with two groups and g-power value of 80%, and st. Voting Classifier (94 percent) outperformed Support Vector Regression in a statistical examination of blood-sample data for predicting diabetes prevalence by gender (80 percent). In summary, the findings of this study demonstrate that the Naive Bayes method is more successful than the SVM algorithm constructed in Python. [ABSTRACT FROM AUTHOR] |
