Performance and Analysis of Predicting Chronic Disease Using Machine Learning Techniques

Abstract

In the field of biomedical and healthcare communities the accurate prediction plays the major role to find out the risk of the disease in the patient. The only way to overcome with the mortality due to chronic diseases is to predict it earlier so that the disease prevention can be done. Such model is a Patient’s need in which Machine Learning is highly recommendable. But the precise prediction on the basis of symptoms becomes too difficult for doctor. To overcome this problem data mining plays an important role to predict the disease. This study analyzes chronic diseases using machine learning techniques based on a chronic diseases dataset from the UCI machine learning data warehouse. We use Heart disease, Kidney disease, Cancer disease and Diabetes disease datasets, In order to build reliable prediction models for these chronic diseases using data mining techniques. The most relevant features are selected from the dataset for improved accuracy and reduced training time. The system analyzes the symptoms provided by the user as input and gives the probability of the disease as an output Disease Prediction is done by implementing the Logistic Regression. By using logistic regression, random forest and decision tree we are predicting diseases like Diabetes, Heart, Cancer and Kidney. For each chronic disease, diverse models, techniques, and algorithms are used for predicting and analyzing. The paper comprises a conceptual model that integrates the prediction of most common chronic diseases.