David Goll

About

I am a Reinforcement Learning (RL) enthusiast, fascinated by how intelligent behaviour can emerge from simple interaction rules.

Trained as a physicist, specialised in statistical physics and modelling of complex systems, I enjoy applying analytical tools to study emergent phenomena. In my Master’s thesis, I studied the learning dynamics of Multi-Agent Reinforcement Learning, uncovering rich and unexpected behaviours even in minimal systems.

I am now seeking opportunities to apply my expertise in various fields, where I can combine my strong quantitative background with my interest in machine learning.

Education

Master of Science in Physics
Humboldt University of Berlin, Germany (Oct 2021 – Feb 2025)
Thesis: "Analysis of Multi-Agent Reinforcement Learning from a Statistical Physics Perspective"
Supervisor: Dr. Jobst Heitzig (Potsdam Institute for Climate Impact Research)

Bachelor of Science in Physics
Heidelberg University, Germany (Oct 2017 – Jun 2021)
Erasmus Exchange: Universidad Autónoma de Madrid, Spain (Aug 2019 – Jun 2020)

Skills & Interests

Research Areas: Statistical Physics, Multi-Agent Systems, Machine Learning, Data Analysis

Programming: Python (PyTorch, scikit-learn, pandas, MLflow), basic Java, Git, shell scripting

Languages: German (Native), English (C2), Spanish (B2), French (A2)

Research & Publications

Below you can access my academic work:

Fun projects

Asteroid Racer

JavaScript HTML5 Canvas

A simple browser-game where players navigate a spaceship to collect energy cores and avoid asteroids.

Future Development: The goal is to train a RL agent to master the game, comparing its performance to human gameplay.

View Code Play Game

RL Agent for Asteroid Racer

🤖

Work in Progress.

Development of a RL agent trained to play Asteroid Racer.

Planned Python PyTorch RL

Contact

Legal Notice & Privacy