In this Diabetes Prediction project in which we diagnostically predicts whether or not a patient has diabetes, based on certain diagnostic measurements included in the dataset like Pregnancies, Skin Thickness, BMI, Age, Glucose etc. Here, we have implemented various Classification techniques and since the dataset is fairly balanced in terms of 0 and 1 outcomes, the best accuracy which I could come up with was 89.6% using ensemble RANDOM FOREST classifier. We have also try KNN Classfier through which we got 88.3% accuracy
Note: In the dataset some independent variables(diagnostic measurement parameters) have '0' values like Insulin, Blood Pressure(which realistically can't be zero). So, we have tried to replace those '0' values by mean values