Diabetes Classification

The experiment uses a dataset containing columns like age,weight,medical_specialty,time_in_hospital,number_inpatient,and various medicines prescribed to them.First,the data is preprocessed and normalized then it is split into training and test.Then various classification algorithms are used on the data .They are Two class Logistic Regression,Two class Neural Network and Two class Boosted Decision Trees.Permutation Feature Importance is also used in the experiment to eliminate features with minimal importance and hence reduce the dimensionality of the data.CrossValidation is performed too. As it turns out from their AUC score,Neural Network and Boosted Decision Tree perform slightly better than Logistic Regression on this data.