A Comparative And Predictive Analysis of Anemia Disease by Using Different Machine Learning Approaches

Abstract

Pregnant women and children are particularly vulnerable to the global public health crisis that is anaemia. If your red blood cell count becomes too low or if their structure is damaged, you may get anaemia. Client reluctance leading to abstinence, a lack of medical professionals and resources in rural locations, and limited financing for medical testing are some of the practical obstacles associated with the clinical diagnosis of anaemia. A number of machine learning techniques have been created to find anaemia because they are cheaper, easier to use, and don't involve damages like the traditional approach accomplishes. Non-invasive techniques, like using machine learning algorithms, are one way to diagnose or find clinical diseases. These days, it's impossible to ignore the use of these techniques for finding anaemia. The utilization of machine learning models to predict anaemia examined in this work. The techniques used may be divided into three phases: preprocessing the dataset, gathering the dataset, and creating a model to identify anaemia. For this comparative analysis used ELM, CNN, RF, NB, KNN and Boosting based XGBoost and AdaBoost models. A performance of several ML models for a predictive analysis of anaemia disease compared in the study with respect to precision, accuracy, recall, and f1-score measures. A significance of precisely forecasting illnesses in the medical domain emphasized by the study. For treatment and preventative efforts to be effective, accurate and timely anaemia estimation is crucial. A use of ML models for predicting anaemia and enhance illness prevention and treatment is demonstrated by this work on the predictive analysis of anaemia disease. The comparative simulation results show the boosting algorithms obtain 100% results in all parameters for the anemia prediction, in compare to other models. The application of sophisticated algorithms and data processing methods can aid physicians in making precise prognostications and decisions, thereby improving patient outcomes. According to the study's findings, ML algorithms may be used to predict anaemia based on common risk factors, which may eventually assist prevent and manage anaemia in children.