Welcome to my personal profile website. I made this to showcase some of my passion projects, professional experiences and other stuff I am working on.

Scientific Machine Learning for Physics. What does this mean?

I tend to have some issues explaining what I do and what I care about. This section is a small synopsis. I try to describe myself as a scientific machine learning engineer, to be specific in Physics. This means I try to apply AI to improve and empower research and the acedemic process.

Physics was, and still is, a field driven by observation. Kepler, Ohm and Plancks laws were all fits to experimental data, not laws derived from first principles. AI and bigger compute allow us to take this a step further. Finding physical laws in many-dimensional data that no researcher could visualise is one of the many great promises big data processing offers.

Ways that AI can empower physics is for example, but not limited to:

• Discovering ODEs/PDEs/equations from data
• Fast surrogate models to replace expensive simulations
• Physics-Informed Machine Learning, AI that learns more efficiently by embedding Physics into models
• Parameter estimation from data
• Discrepancy modeling to learn “residual” Physics
• Improve (measurement) resolution using ML techniques

I aim to expand research and applications in this direction.

Other (Work/Research) Interests

Though Physics-Informed Machine Learning holds my current interest, there are other related topics I also quite like. These inlcude

• Bayesian Machine Learning & Active Learning
• Predictive Maintenance & Digital Twinning
• Reinforcment Learning

As you may notice, these all remain in the field of ML/Data Science or engineering.