Hybrid Algorithm for Security of Data Over Wireless Transfer with Disease Prediction in Machine Learning

Data security is major issue in the entire field. With the evolving technological advancements in our day to day life security of personal data is a primary concern for all. Many intruders and hackers are eagerly waiting for personal data of the people which can be used in various means for their works. They always have an eye on the personal information of the people. And the second issue is that diagnosing of diabetes by the doctor will be some time prone to mistakes because it is natural that human makes mistakes but not the machines. So to overcome this issue we develop a hybrid secure algorithm which encrypts the patient personal information which they wish not to share to anyone outside the hospital by crypto AES encryption and hashing using SHA-256 and then storing the data in database which will be accessible only to the patient and the doctor. Together with encryption of the patient personal data and medical data our proposed system will effectively evaluate the data by a trained machine learning model to check whether a person is actually suffering from diabetes or not. We use logistic regression for evaluating the patient medical data to predict the health status so mistakes can be avoided in prediction as the prediction is done by the trained machine. Then doctor can give the treatment depending on the health status. As a result this system will give a sustainable solution to people for safely transmitting their personal data as well as medical data through wireless medium along with diabetes disease prediction.