Curriculum Vitæ

Education

• Ph.D in Time Series Anomaly Detection (expected): Conducting research on deep learning for time series, with focus on anomaly detection, forecasting, and the development of foundation models for zero-shot forecasting.

• M.S. in Mathematics & Artificial Intelligence, Paris-Saclay University, 2022-2024. Coursework: Machine Learning (supervised, unsupervised, deep learning, reinforcement learning), Statistical Learning, High-Dimensional Modeling, Graphical Models, Optimization (incl. for CV), Big Data Systems (SQL, HDFS, MongoDB), Online Learning, Conformal Prediction, Data Challenge.

• B.S. in Mathematics, Paris-Saclay University, 2019-2022

Work Experience

• April 2024 - September 2024: A.I. Research Intern
  • THALES
  • Training-Free Metric for Neural Architecture Search
    Conducted a research-oriented internship exploring the Neural Tangent Kernel as a training-free metric for Neural Architecture Search. Additionally, I investigated and compared various search strategies to optimize neural network performance effectively.
• March 2023 - August 2023: A.I. Engineer Intern
  • AZAP
  • Uncertainty Analysis in a Machine Learning Forecasting Approach
    Developed and implemented a machine learning model using Gradient Boosting to predict sales of promoted products. The project emphasized uncertainty analysis to enhance forecasting reliability.
• 2021 - Present: Math Tutor
  • France
  • Private Mathematics Tutor
    Provide personalized mathematics classes to students, helping them strengthen their understanding and improve their academic performance.

Publications

PatchTrAD: A Patch-Based Transformer focusing on Patch-Wise Reconstruction Error for Time Series Anomaly Detection

Samy-Melwan Vilhes, GILLES Gasso, Mokhtar Z. Alaya (2025). "PatchTrAD: A Patch-Based Transformer focusing on Patch-Wise Reconstruction Error for Time Series Anomaly Detection." EUSIPCO 2025.

A control of the false anomaly rate for Anomaly Detections problems on Images

VILHES Samy. (2024). "A control of the false anomaly rate for Anomaly Detections problems on Images."Open Science.

Understanding Neural Tangent Kernel: Key Theories and Experimental Insights

VILHES Samy. (2024). "Understanding Neural Tangent Kernel: Key Theories and Experimental Insights." HAL Open Science.

The Theory behinds Variational Auto-Encoders

VILHES Samy. (2023). "The Theory behinds Variational Auto-Encoders."Open Science.

The Theory behinds Denoising Diffusion Probabilistic Models

VILHES Samy. (2024). "The Theory behinds Denoising Diffusion Probabilistic Models."Open Science.