Reinforcement learning
Pokémon: a reinforcement learning adventure
Classic reinforcement learning, told as a retro Pokémon battle.
Senior Scientist · Lecturer · ETH Zürich
Carlos Cotrini
I teach machine learning and computer science, and research how we learn them.
I am a Senior Scientist and Lecturer at the Institute of Machine Learning, ETH Zürich, under the supervision of Professor Emeritus Joachim Buhmann. Teaching is my focus: I teach machine learning and computer science to audiences from master's students to managers in continuing education, and my research now turns to computer-science education itself.
ccarlos (at) inf.ethz.ch · CAB, ETH Zürich
News
- 2026 Our paper Transforming Confusion into Diffusion, on teaching machine learning bottom-up and from scratch, received the CER Best Paper (Global) award at the ACM SIGCSE Technical Symposium 2026.
- Spring 2026 Five new interactive visualizations for my courses: PCA on 175 years of Senate votes, eigenfaces, neural networks, a sunrise-painting VAE, and reinforcement learning as a Pokémon battle.
- Autumn 2025 Two recorded lectures from Advanced Machine Learning are online with full notes: regression from least squares to neural tangent kernels, and a recap on convex optimization.
From the lecture hall
Featured lectures
Regression, from least squares to neural tangent kernels
Convex optimization with tridents and lightsabers
Machine learning, made tangible
Interactive visualizations
Generative models
Sunrise Machine
A variational autoencoder that learns to paint sunrises, and lets you wander its latent space.
Principal component analysis
The Senate, projected
A data story that watches 175 years of U.S. Senate roll-call votes collapse onto two principal components.
About
I am a lecturer at the Institute of Machine Learning at ETH Zürich, under the supervision of Professor Emeritus Joachim Buhmann. I teach courses in machine learning and computer science with a focus on making complex concepts accessible to students and bridging theory with practical applications.
My research journey includes a PhD in information security under the supervision of Prof. David Basin, where I developed expertise in access-control policy mining and the privacy and security analysis of web technologies. Earlier, I worked on infon logic under the guidance of Prof. Yuri Gurevich.
Today my work focuses on privacy-preserving and robust machine learning, large-scale security and compliance analysis, and the craft of teaching these subjects well, through intuitive explanations and from-scratch implementations that demystify how modern models actually work.