Diabetes Prediction Using Blood Sample Data with Novel Voting Classifier over Random Forest

This study focuses on how to predict diabetes using blood sample data and machine learning algorithms like the Voting Classifier over Random Forest technique. The proposed prediction models were trained evaluated a dataset that included seven variables: glucose level, diastolic pressure, thickness, insulin levels, BMI, age, skin. new classifier (VC) (RF) are used of 1495 records with 10 features, size=5, two groups g-power value 80%. With threshold 0.05, confidence interval 95 percent, standard deviation one deviation, patients’ information was acquired from variety websites. framework built VC algorithm, resulting in successful research (95%) technique (85 percent). percent interval, two-tailed t-test revealed statistical significance 0.001 (p0.05). shows algorithm’s results more accurate than RF approach, which written Python.