Curability Prediction Model for Anemia Using Machine Learning

Abstract

Anemia, defined by the WHO as an insufficient red blood cell count is the most common blood illness globally. This illness affects one's condition of life as a disease and as well as a symptom. In terms of patient therapy, it's critical to get a proper diagnosis of the type of anaemia. The rising count in patients and hospital priorities, as well as the difficulty in obtaining medical specialists, could make such a diagnosis challenging[2]. The current study provides a technique for detecting anaemia in clinical settings. We use CBC (complete blood count) data obtained from pathology centres to examine supervised machine learning techniques - Naive Bayes, LR, LASSO, and ES algorithm for prediction of anaemia[1]. And make a assumption of the patients, wheather He/She get Cured or not Cured after 90 Days. In comparison to LR, LASSO AND ES, the results suggest that the NaiveBayes approach outperforms in terms of accuracy