Ensemble Learning on Diabetes Data Set and Early Diabetes Prediction

Ensemble Learning is a process through which multiple classifiers are used to find solution to a problem. With the use of Ensemble Learning on a Diabetes data set a model is trained to predict the early onset of diabetes. This model targets for two major objectives: whether the person has a risk of diabetes in the near future or not and the risk probability of having diabetes associated with the person. The dataset consists of a basic set of questionnaires which would be helpful in training the model based on various classifiers or experts that strategically train the model. The libraries used in the system are Scikit-Learn, Pandas and Numpy. Once the model is trained it requires no retraining and could be helpful in predicting the early onset of diabetes in a person based on a basic questionnaire on the lifestyle of the subject such as regular lifestyle habits, eating habits, family history and medical history. Since predicting diabetes before it occurs is a problem complex to solve as the disease is basically a lifestyle disease which requires self-assessment of the patient and prevention of the disease at every step, so for reducing the uncertainty ensemble learning is used for prediction of diabetes risk in future as ensemble learning is a collection of various machine learning models.