Teaching

My teaching connects probability, statistics, machine learning, and computational practice.

I have designed and delivered graduate-level courses, developed exercise and laboratory material, contributed to assessment and examination, and supervised academic and industrial student projects. My course design emphasizes the interaction between mathematical tools, current research papers, and reproducible computation.

Courses designed and taught

  1. 2021

    High-Dimensional Probability and Statistics for Data Science

    KTH Royal Institute of Technology

    Designed the syllabus and delivered a fourteen-meeting graduate course. Each meeting paired a probabilistic or statistical tool with a research paper, connecting concentration, random matrices, empirical processes, and high-dimensional inference with modern applications.

    14 × 2-hour meetings Original syllabus Research-paper discussions
  2. 2020

    Statistical Machine Learning

    KTH Royal Institute of Technology

    Designed the syllabus and delivered an eight-meeting course covering statistical foundations and core machine learning methodology, with emphasis on the assumptions, guarantees, and computational behaviour of the methods.

    8 × 1-hour meetings Original syllabus Statistics and computation

Course instruction and materials

  1. 2018–2020

    EL2820 Modelling of Dynamical Systems

    KTH Royal Institute of Technology

    Led exercise sessions and laboratory work, graded assignments and examinations, and contributed to course assessment. I co-authored and revised the 2019 exercise booklet and developed a Jupyter-based homework assignment connecting system identification theory with numerical experimentation.

    Exercises and laboratories Grading and examinations Course booklet Jupyter homework

Teaching practice

Mathematics and current research

Course meetings connect foundational results with selected papers so that students see how abstract tools are used in active research.

Computation as part of reasoning

Numerical experiments and notebooks are used to test assumptions, visualize behaviour, and complement analytical arguments.

Flexible course delivery

During the 2020 COVID disruption, I recorded my own exercise sessions and provided hybrid support for students working remotely.

Student supervision

Bachelor’s thesis supervision

  • 2020 Daniel Hirsch and Tim Steinholtz KTH Royal Institute of Technology
  • 2019 Emil Backman and David Petersson KTH Royal Institute of Technology
  • 2019 David Mainwaring and Jonathan Österberg KTH Royal Institute of Technology
  • 2018 Mikaela Åstrand and Oscar Bergqvist KTH Royal Institute of Technology

Master’s and research supervision

  • 2023–2024

    Archith Athrey — learning-based stochastic control

    Co-supervised his master’s research at TU Delft, contributing to problem formulation, theoretical development, methodology, and analysis. The collaboration continued beyond the thesis and led to a joint publication at the 2024 European Control Conference.

  • 2018–2019

    Carl-Johan Larsson and Fredrik Håkansson — industrial master’s thesis

    Supervised the thesis User-Based Predictive Caching of Streaming Media in connection with Spotify. I primarily mentored the machine-learning component, including recurrent-neural- network modelling, trajectory prediction, and evaluation using large-scale mobility data.

  • 2025–2026

    Alexandre Alouadi, PhD candidate

    Research collaborator on two projects forming part of his doctoral research. The collaboration includes problem formulation, theoretical development, methodology, experimental design, and preparation of joint manuscripts.