Hybrid Machine Learning Model for Chronical Disease Prediction

Abstract

The existing methods for predicting chronic diseases have a number of issues with feature extraction, outlier removal, and classification. The primary goal of this research project is to address current shortcomings. The prediction of chronic diseases will be more accurate once the current flaws are eliminated. As a result, it is critical to analyse the changes in chronic diseases using epidemiological data and to forecast the disease using socioeconomic data.  In this study, a hybrid model based on CNN and Bidirectional LSTM is given. For feature extraction, CNN is utilized, while logistic regression is employed for classification. Using CNN, the collected features were fed into a bidirectional LSTM for classification. The suggested framework is put into practice using Python, and the accuracy, precision, and recall of the output are examined.