ML Based Interactive Disease Prediction Model

The application of Machine learning algorithms to predict diseases is one of the finest methodology to reduce heavy work load on doctors and related medical staff. Based on the World Health Organization (WHO) report, about 85% heart disease deaths are due to Heart Attacks and Heart Strokes. In India the average death rate due to cardiovascular diseases is about 272 per 10,000 population which is greater than global average of 235 per 10,000 population. From the recent survey results, which was released by the Union Ministry of Family and Health Welfare (MoFHW), the Diabetes disease positive ratio is gradually increasing in India. 11.5 percent people were tested positive for Diabetes among urban and rural Indians who are with age 45 and above. Even there is availability of wide range of treatment methods of heart stroke patients & diabetes, Heart attack with Diabetes is the major cause of death in all parts of rural and urban areas of entire India. There are several factors causing heart and diabetes problems which include Age, Gender, Blood Pressure, Glucose levels, Skin thickness and Insulin. These are easily measured in primary care facility centres. The accurate estimation and analysis of heart & diabetes disease patients reports data may help in predicting future heart problems including diabetes. Globally, the application of computerized machine learning methods to predict future problems is in trend now. The Health Monitoring Departments and Fields uses machine learning algorithms to predict and analyse in a wider way to solve problems in fraction of seconds. From the famous proverb “Prevention is Better Than Cure”, if we apply this to medico and health field we can save people from major Heart Diseases (HD's) along with Diabetes. The proposed Dual disease prediction technique is user interactive based method. The proposed method observe inputs from the end user with realistic data to predict heart and diabetes disease. In the presented work, we used Logistic regression model (LR) and Support vector machine (SVM) model for prediction of diseases. The proposed model works with 85 and 78 percent accuracy in prediction of heart and diabetes diseases respectively.