Logistic Regression Analysis for the Classification of Human Pulse Data

Abstract

For its simple, effective and non-invasive approach, radial pulse signals have been used for diagnosing the health condition of humans. Despite the fact that there are numerous methods for acquiring pulse data and data processing techniques, no model based on predictions and classification is still to be constructed. Hence, the suggested work intended to classify human pulse signals using a logistic regression model that is better at understanding them. Logistic regression was used to evaluate and extract dependent and independent variables from a dataset of human pulses. This method identifies the best-fitting model to represent the relationship between dependent and independent variables, as well as the performance of each dosha, such as Vata, Pitta, & Kapha that enables medical practitioners/clinicians in understanding the relationship between each dosha, determining the performance of each dosha will assist him in finding the dominance among three and, based on this, he can recommend treatment.