Hussein El Amouri

Hussein El Amouri

Ph.D, Assistant Research of Artificial Intelligence

Strasbourg University

Biography

I am currently a fellow researcher working on active learning for constrained proposition to facilitate incremental and constrained clustering applied to remote sensing and time series data. This work extends the research conducted during my three-year Ph.D., which introduced the model called Constrained DTW Preserving Shapelets (CDPS) and Shapelet Cluster Explanation (SCE). The objective of the current work is to leverage SCE and CDPS to propose constraints actively.

Interests
  • Active Learning
  • Incremental Learning
  • Contrastive Learning
  • Constrained Clustering
  • Time series
Education

  • PhD in Computer Science, 2023

    University of Strasbourg

  • Master in Signal and Image Processing, 2020

    Grenoble INP Phelma

  • Telecomunication and Electrical Engineering, 2020

    Lebanese University Faculty of Engineering

Experience

 
 
 
 
 
ICube Laboratory
Post-Doc (researcher assistant)
ICube Laboratory
October 2023 – Present Illkirch-Grafenstadden

Responsibilities include:

  • Investigating different approaches for constraint proposition in constrained clustering. Based on works from active learning and pseudo-labeling
  • Study the incremental addition of constraints to CDPS: Constrained DTW preserving Shapelets, and the effect on remote sensing data.
  • Deploy an open-source code of the findings.
 
 
 
 
 
University of Strasbourg
Ph.D. Candidate
University of Strasbourg
October 2020 – September 2023 Strasbourg
Investigate the different ways to enhance clustering results for remote sensing data and techniques to explain the clustering output. This work settled on using shapelets to learn speech subsequences of the time series in the datasets that can differ between the different possible samples in the dataset. ….
 
 
 
 
 
University of Strasbourg
lecturer
University of Strasbourg
October 2021 – Present Strasbourg
 
 
 
 
 
GIPSA Laboratroy, Grenoble INP-Phelma University
Machine Learning Intern
GIPSA Laboratroy, Grenoble INP-Phelma University
March 2020 – September 2021 Grenoble

Publications

Filter publications
(2023). Constrained DTW preserving shapelets for explainable time-series clustering. Pattern Recongintion.

PDF Cite Code Dataset DOI

(2022). Constrained DTW Preserving Shapelets. In ECML.

PDF Cite Code Dataset DOI

(2020). Evaluating the Extrapolation Capabilities of Neural Vocoders to Extreme Pitch Values.

PDF Cite DOI

Contact

For inquiries and detailed information, feel free to reach out. I look forward to connecting with you.