I am Mert Dil, a Physics M.Sc. student at Technical University of Munich (TUM) working at the intersection of physics, data science, and machine learning.
My work focuses on applying computational and data-driven methods to complex physical systems, especially in energy and plasma research.
I currently contribute to research projects in fusion energy and battery systems, combining physics-based modelling with machine learning approaches.
- ⚡ Machine learning for plasma heating and diagnostics
- 🔋 Hybrid ML–physics models for battery degradation
- 🧠 Deep learning for scientific data analysis
- 📊 Scientific computing and simulation pipelines
-
🎓 M.Sc. Physics – Technical University of Munich (TUM)
Focus: computational physics, machine learning, and energy systems. -
🧪 Max Planck Institute for Plasma Physics (IPP)
Working on machine-learning methods for ion cyclotron resonance heating (ICRH) diagnostics. -
🚀 German Aerospace Center (DLR)
Developed hybrid ML models (LSTM + physics constraints) for battery degradation forecasting. -
🔭 CERN CMS Research Group – ITU
Applied generative deep learning (WGAN-GP) to accelerate particle physics simulations.
Predictive modeling of solar-wind and heliospheric dynamics using LSTM networks.
Tech stack
Python • Keras • Pandas • Scikit-learn
Highlights
- Modeled plasma flow variability in solar wind data
- Built an end-to-end deep learning pipeline for time series forecasting
🔗
https://github.com/Mertdil/Forecast-Time-Series-using-Space-Enviroment-
Deep learning based image classification and object recognition.
Tech stack
Python • TensorFlow • OpenCV
Highlights
- Built a full preprocessing and training pipeline
- Applied CNN architectures for multi-class classification
🔭 Currently exploring research directions in
- AI for Plasma Physics
- Energy Systems & Battery Modeling
- Scientific Machine Learning
- Physics-driven simulations