SYMPTOMATIC ASSESSMENT OF DISEASES USING DECISION TREES AND ANALYSIS OF ELECTRONIC MEDICAL RECORDS

Abstract

Supervised machine learning algorithms have emerged as the primary data mining tool.
The use of health data to diagnose disease has lately revealed the potential use of these technologies.
The purpose of this research is to find various forms of regulated machine learning algorithms as well
as major trends in measuring performance and illness risk. In this article, we will attempt to anticipate
patient illnesses based on their symptoms. We employ the decision tree algorithm to reach this aim, which
will aid in the diagnosis of patients' health. The data set includes physiological measures for 42 different
illnesses (diseases) and 129 different features (symptoms). We created a categorized decision tree model
that uses standardization techniques known as format reduction to generalize data and delivers training
to a dataset in a short amount of time. Developed trained models are then utilized to forecast illnesses,
including their causes and preventative strategies, after they have been normalized.