AI & Physics-Informed Neural Networks (PINNs) Workshop

Bridging Artificial Intelligence and Mathematical Modeling through Deep Learning & PINNs.

As part of the First International Conference on Advances in Operator Theory and Applications (ICAOTA'2026), this immersive workshop offers a cutting-edge journey into the fusion of AI and differential equations. Step into the future of computational science.

Workshop Details

  • 📍 Location: Hammamet, Tunisia
  • 🗓️ Dates: March 24–25–26, 2026
  • ⏱️ Duration: 5 hours
  • 👨‍🏫 Instructor: Pr. Bassem Ben Hamed
  • 📧 Contact: bassem.benhamed@enetcom.usf.tn

Workshop Agenda

Session 1: Foundations and First Implementation (2.5h)
  • Theory of PINNs & Universal Approximation Theorem
  • PINN architecture & loss function design
  • Lab: Solve a second-order ODE – Harmonic Oscillator
Session 2: PDEs and Real-World Applications (2.5h)
  • PINNs for PDEs and inverse problems
  • Boundary/initial condition enforcement
  • Lab: Solve the 1D Diffusion Equation using PyTorch/TensorFlow

Learning Outcomes

  • Master the theory and architecture of PINNs
  • Build models that embed physical constraints
  • Gain experience with TensorFlow/PyTorch in scientific computing
  • Bridge AI and applied mathematics in real-world scenarios

Tools & Technologies

You'll be working with:

  • Python 3.x • TensorFlow or PyTorch
  • Jupyter Notebooks • NumPy/SciPy

Who Should Attend?

This workshop is ideal for:

  • Graduate students and postdoctoral researchers
  • Professors and scientific programmers
  • Engineers and data scientists in modeling/simulation

Requirements

  • Basic linear algebra and differential equations
  • Familiarity with Python programming
  • Curiosity and a passion for AI-driven science!

Certification

Participants will receive a signed Certificate of Participation upon successful completion.

Registration

Participation is free for ICAOTA’2026 registrants.