Inverse problems
Estimating latent quantities and spatially varying fields when direct measurements are incomplete or expensive.
Profile
Research overview.
My research focuses on scientific machine learning and inverse problems. I am interested in parameter estimation, anomaly detection, and learning from limited or irregular observations.
Much of the work uses physical structure and governing equations to make models more reliable and easier to interpret in monitoring settings.
Scientific machine learning
Designing models that respect governing equations, structure, and known regularities instead of relying on brute-force pattern matching.
Monitoring systems
Working on methods that can detect change, localize anomalies, and support decisions from limited sensing.
Irregular domains
Building approaches that stay stable when geometry, discretization, or data density are uneven.
Papers
Recent papers.
Mechanics-Informed Anomaly Detection
A mechanics-informed autoencoder for structural anomaly detection that uses physical structure to improve reconstruction and interpretability.
ParaFIND
A framework for parameter-field inference on non-uniform domains, aimed at recovering hidden quantities from sparse and irregular observations.
Neptune
A model for estimating parameter fields in multi-physics PDE settings, combining structure from the equations with learned representations.
Research
Current research areas.
Parameter estimation from limited observations
- Estimating latent variables and spatially varying fields from partial measurements.
- Working on methods that remain stable when observations are sparse or uneven.
Physics-informed learning
- Using mechanics and PDE structure to guide model design and training.
- Combining learned representations with known physical constraints.
Monitoring and anomaly detection
- Developing methods for detection, localization, and interpretation.
- Focusing on settings where sensing is noisy, incomplete, or expensive.
Trajectory
Academic background.
Ph.D. research
Michigan State University
- Research in scientific machine learning, inverse problems, and monitoring.
- Affiliated with the Hybrid Analytics Lab.
Engineering training
Ecole Polytechnique de Tunisie
- Training in engineering, mathematics, and computation.
- Early focus on physical systems, modeling, and data-driven methods.