Machine Learning and Data Mining

This research domain delves into the integration of machine learning and artificial intelligence in various sectors, focusing on medical diagnosis, road quality and complex network analysis.

The analysis of complex networks through data mining methodologies, particularly with the “Reduced Google Matrix – REGOMAX” approach. Investigating various types of networks, the analysis aims to comprehend the underlying mathematical properties that govern these intricate systems. The ongoing research involves ameliorating the algorithm by dealing with various network types and combining neural networks with machine learning in the decision-making step of sensitivity analysis. This work is notably scarce worldwide, signifying a pioneering effort in this field.

Additionally, a complementary area within this research domain involves developing an ontology using machine learning algorithms. The project focuses on text comprehension, classification, relationship establishment between text categories, and automated generation of new categories. This work contributes to assisting company directors in understanding communication processes efficiently.

The research extends to collaborating with renowned medical institutions such as Centre de références des maladies rares orales et dentaires, Hôpitaux universitaires de Strasbourg, O-Rares network, the IGBMC, CERBM-gie and the Fondation Maladies Rares, the ongoing project “D-IA-GNO-DENT” addresses early detection and prevention of rare oral and dental diseases. Leveraging machine learning algorithms and AI techniques, this initiative advances image processing, computer vision, and segmentation for precise identification of anomalies, contributing to improved medical diagnostics.

Simultaneously, the study embraces the development of AI applications for infrastructure maintenance and governance in transportation systems. Under the project “Maintenance through Imaging in Artificial Intelligence for Safe Roads & Augmented Gouvernance – MIIRAG“, the focus is on creating a collaborative artificial intelligence platform for the detection of cracks on pavements using vision-based AI and reinforcement learning. The aim is to enhance accuracy, confidence, and transparency in road maintenance operations, making use of IoT devices and autonomous vehicles.

The amalgamation of machine learning and data mining applications not only contributes to advancements in medical diagnostics, infrastructure maintenance, and complex network analysis but also reflects a pioneering effort in various interdisciplinary domains.