System Modeling and Prediction of Respiratory Diseases using Machine Learning

Abstract

In the first part of this paper a human respiratory model and their responses are presented and in the second part machine learning algorithm is used to predict the respiratory diseases from the datasets. This endeavor proposes a simple representation of the mathematical model of the human respiratory system comprises of nasal cavity, trachea, bronchi and alveolar sacs. Ranging from a very common disease called bronchitis to highly perilous diseases like emphysema are considered in this venture and the variation of responses are observed with input pressure. In the conventional computation and control the system modeling is very much essential for analysis, whereas, machine learning algorithms use computational methods to learn information directly from data without relying on a predetermined equation as a model. Logistic regression algorithm is applied to find natural patterns in respiratory data that generate insight and help to make better decision and prediction of respiratory diseases. The Logistic regression model makes a clear decision boundary between the respiratory diseases. Keywords: Bronchitis, Emphysema, Logistic regression, Mathematical model of the human respiratory system.