teaching

Practical machine-learning material developed while assisting Dr Sathiskumar Ponnusami at Queen Mary University of London.

I wrote these tutorials for my courses in mechanics and scientific machine learning.

teaching footprint

A collection of materials I maintain for my courses.

Course-facing focus Every tutorial is a runnable notebook with short explanations and follow-up examples. They work just as well for self-study as they do in the classroom.
Ownership and upkeep I write and maintain all the tutorials here myself. You can check the commit history if you want to see how they evolved.
Scope of material The tutorials focus on the exact tools I use every day: phase-field mechanics, ML modelling, and physics-informed neural networks.

The main tutorial categories.

PyTorch Tutorials Hands-on model-training workflows from basic loops to applied physics-informed examples.
Machine Learning Fundamentals Regression, classification, model diagnostics, and generalisation habits for a clean starting point.
Scientific ML Mechanics-aware ML applications that focus on interpretation and model behavior under constrained engineering data.

Feel free to use these in your own teaching or reference them in your work.

Core tutorial stack A sequence of five tutorials taking you from basic PyTorch loops all the way to training your first PINN.
Notebook-first delivery Code comes first. You can run the cells, see the intermediate outputs, and break things to see what happens.
Publicly cited references I link back to standard engineering textbooks and ML primers so you know exactly where the math comes from.

For educators Feel free to borrow these notebooks for your own classes.
For collaborators Pick a path depending on what your lab needs: PyTorch basics, ML foundations, or scientific ML.
For students Work through the notebooks from top to bottom. They start with the basics and end with applied mechanics.