Diabetes Prediction using Logistic Regression and Feature Normalization

Diabetes is one of the many major issues in medical field and lakhs of people are affected due to this diabetes. From many years many researches are going on this problem to detect this diabetes. Here we are mainly concerned towards women because during pregnancy they may get diabetes which is also termed as gestational diabetes and due to this there is a higher chance of getting diabetes called type2 in future and this occurs when our human body doesn't use the insulin hormone and it is unable to prepare it. Therefore many methods are there in literature that is used to classify whether a particular human being gets diabetes in future or not. Generally the dataset used for this purpose is Pima Indian diabetes dataset and it is mainly used by the researchers to classify whether an instance has diabetes or not. There are a lot of problems if this diabetes is not treated and it may leads to other organ related diseases. The main problems occur to kidneys, eyes and heart etc. the normal method that is used for this diabetes detection is to visit a hospital or any health care center and we have to reach doctor for treatment. Many researches in machine learning are going on for this purpose and many methods are proposed using the data of people of past and tries to develop models that is used to predict diabetes. In this we are going to propose a method using logistic regression which is technique that is used for detection of diabetes.