Résumé:
Early detection of heart disease is a crucial factor in its successful management. In
recent years, the medical field has seen the emergence of various methods based on ma-
chine learning and deep learning to predict heart disease before it occurs. Indeed, heart
disease remains a major cause of death worldwide. It is therefore essential to diagnose and
treat them as quickly as possible. The aim of this work is to develop a heart disease pre-
diction system capable of detecting the early onset of heart disease, which can prove fatal.
This research project’s approach aims to detect heart disease using supervised ma-
chine learning techniques. To this end, we explored several supervised machine learning
algorithms in order to identify the most efficient model for achieving the optimum results.
Specifically, we focused on how to apply our proposed model approach to the early de-
tection of heart disease, which may help to improve pre-diagnosis and reduce the risks of
complications of heart disease.
Our work focused on improving the decision tree (DT) algorithm for heart disease
detection, which was already considered the best algorithm in this field. In addition, we
proposed specific improvements to this algorithm to increase its efficiency. Using the same
dataset as other supervised machine learning algorithms, we were able to demonstrate that
our improved algorithm produced better results. These results reinforce the idea that our
approach offers a more effective solution for heart disease detection than other existing
algorithms.
