Lazhar LABIOD : soutenance HDR

Lazhar LABIOD

Unsupervised Data Embedding and Co-clustering Beyond Sequential Learning

Abstract

This HDR manuscript presents most of the work I have carried out since my arrival in 2013 within the MLDS team at LIPADE and since 2020 within the Borelli center, as a teacher-researcher in computer science. Until then my research had been part of unsupervised learning and data science with the development of efficient and effective algorithms to explore and summarize increasingly complex data. As well as the implementation of exploratory models and tools to help in the exploration and analysis of complex data. (see for example :TensorClust and CAEclust). The issues related to textual, biomedical, social networks, attributed graphs and tensor data gradually led me to focus my research work in the context of unsupervised learning, namely dimension reduction, clustering, and coclustering of complex data. More specifically, I was interested in the simultaneous learning of dimension reduction and co-clustering, for the processing and analysis of complex data.

My research has also been an essential mine for the design and animation of my lessons in machine learning, data science, and dimension reduction as part of the Masters M1 and M2 MLDS and AMSD since 2014 at Ufr math info and at the Ufr Science Fondamentale et Biomédicale, Université. Paris Cité.