Diabetes predictor is a simple app that determines whether a female patient has diabetes based on certain parameters. This app uses PIMA Indians Diabetes Dataset as the training data. The underlying ML model used by this app is a Scikit-Learn Logistic Regressor. This algorithm marginally edged over others like Support Vector Machine and Random Forest Classifier. It is packed inside a Scikit-Learn pipeline, which has a column transformer at its front with a Standard Scaler object inside it to normalize the input data and pass it to the ML model.
Pregnancies:- The number of children this patient has given birth to in her life so far. Min value:- 0, Max Value:- 17
Insulin:- The amount of microunits of Insulin per milliliter of blood (mcU/mL). Min value:- 0.0, Max Value:- 846.0
Blood Glucose:- The amount of milligrams of Glucose per decilitre of blood (mg/dL). Min value:- 0.0, Max Value:- 199
Age:- The age of the patient in years. Min value:- 21.0, Max Value:- 81.0
Skin Thickness:- The skin thickness of the patient in millimeter(mm). Min value:- 0.0, Max Value:- 99.0
Blood Pressure:- The Blood pressure of the patient in millimeters of mercury (mmHg). Min value:- 0.0, Max Value:- 122.0
Pedigere Function:- The Pedigere Function of the patient from her medical history. Min value:- 0.078 Max Value:- 2.42