Approche de prédiction par télésurveillance à base de Data Mining.

Abstract

Following the technological evolution, in particular the mobile approach, scientific research has been oriented towards the exploitation of these advances for remote predictive decision support. A major interest of researchers has had a great impact in the medical field because of its very positive influence for the care of the patient aimed at its assistance and the reduction of cases of death due to follow-up and the problem of time of treatment. emergency action. This is how telemedicine has become an issue of great importance, it is based on the manipulation and analysis of a large volume of medical data. The aim of this thesis is firstly to exploit a new approach to data analysis, namely Symbiotic Organisms Search (SOS) for Data Mining for data classification, and secondly, to propose improvements to this metaheuristic. This improvement relies on the integration of speed in SOS as a new parameter to explore the search space efficiently and avoiding premature convergence. We also develop a conceptual and practical architecture for applied telemedicine for decision support for the knowledge of the type of breast cancer (benign or malignant). This study allowed us to achieve excellent results and findings in terms of data classification.