Diagnostic Approach To Anemia In Adults Using Machine Learning

Abstract

Computer-aided illness diagnosis is less expensive, saves time, is more accurate, and removes the need for additional personnel in medical decision making. Many nutrition surveys indicate that about a quarter of the world's population is anaemic. As a result, there is a pressing need to create an effective machine learning regressor capable of properly detecting anaemia. The goal is to find out which individual classifier or group of classifier combinations obtain the highest accuracy in red blood cell categorization for anaemia detection. We used Lasso and Ridge regressions to detect and estimate the anaemia. However, the classifier Ridge performs better achieves an accuracy higher than the Lasso regression. Hence to achieve maximum accuracy in medical decision making, a better and powerful algorithm should be used. The outcomes of this algorithms decides whether the patient is infected with anaemia or not. The proposed version generates a better response to the inputs to confirm the disease.