I am currently a data scientist at Veolia Eau d'Ile-de-France solving problems related to the production and distribution of drinking water.
I have completed my PhD thesis in change-point detection and data analysis at Ifsttar / Grettia and Veolia Environnement Research & Innovation. This PhD work was done under the supervision of Patrice Aknin and Allou Samé from Ifsttar.
My research interests mainly involve industrial diagnosis, change-point detection, sequential hypothesis testing, segmentation of functional data, finite mixture models and EM-like algorithms in unsupervised and semi-supervised learning.
Thesis: Detection and classification of temporal CAN signatures to support maintenance of public transportation vehicle subsystems
- 2016/11: Presentation of two papers
- Modeling and Clustering Water Demand Patterns from Real-World Smart Meter Data
- ResiWater: A Franco-German Project for Augmented Resilience of Water Distribution Systems following Severe Abnormal Events
- 2016/10: Presentation of the paper Segmenting Multivariate Time Series of Water Flow: a Prior Tool for Contamination Warning Systems at the International Conference on Embedded Systems in Telecommunications and Instrumentation (ICESTI 2016) organized by the University of Annaba (AL) from the 24th - 26th October 2016.
- 2016/01: Presentation of the paper Décomposition et classification de données fonctionnelles pour l'analyse de la consommation d'eau at the workshop Clustering and Co-clustering (CluCo 2016) at the french conference Extraction et Gestion des Connaissances (EGC) (EGC 2016) organized by the University of Reims (FR) from the 18th - 22nd January 2015.
- 2015/12: Presentation of the paper Forecasting Quality of Raw Water to Optimize Drinking Water Production at the International Conference Water, Megacities and Global Change (Eaumega 2015) organized by ARCEAU IdF and devoted to the theme of water in megacities. It will be hosted by UNESCO in its Paris HQ, 2015 - December 1st to 4th, during the first week of the COP21.
- 2015/10: Presentation of the poster Décomposition et clustering de données fonctionnelles pour l'analyse de la consommation d'eau potable (french) at the Clustering day workshop (Clustering 2015) hosted by Orange Labs in Issy Les Moulineaux (FR) on the October 20th 2015.
- 2015/09: Presentation of the paper An Incremental Sensor Placement Optimization in a Large Real-World Water System at the 13th Computer Control for Water Industry (CCWI2015) conference (CCWI 2015) organized by De Montfort University in collaboration with the University of Exeter and the University of Sheffield. It takes place in Leicester (UK) from the 2th - 4th September 2015.