Advanced Deep Learning

Advanced Deep Learning (BEV033DLA) covers the algorithmic and theoretical foundations of deep neural networks, in the Open Informatics master’s program at CTU in Prague.

Term: Spring semester

Overview

Advanced Deep Learning (BEV033DLA) introduces deep neural networks and Deep Learning, a branch of Machine Learning and Artificial Intelligence. It provides the algorithmic and theoretical concepts needed to design and train neural networks successfully, while building technical and practical skills in the domain.

The course is best suited to master’s students, who are expected to arrive with basic knowledge of Machine Learning and Artificial Intelligence.

I have contributed as one of the lab instructors since 2026, supporting the practical and theoretical tutorials.

I ran the lab on Vision Transformers, and designed and ran the final lab as a student hackathon, which turned out to be a favourite with the class.

Study Programs

The course belongs to the Open Informatics follow-up master’s program at CTU, appearing in two of its specializations:

  • Artificial Intelligence
  • Computer Vision

It is taught in English, supervised by the Department of Cybernetics, and carries 6 ECTS.

Teaching Team

Lecturers: Alexander Shekhovtsov and Giorgos Tolias.

Lab instructors: Vladan Stojnić, James Hooper, Bill Psomas, Giorgos Kordopatis-Zilos, Andrii Yermakov, Jan Šochman, and Tilemachos Aravanis.

Format

The course runs in a 2P+2C format: weekly lectures, with practical and theoretical labs alternating each week. Practical labs work through homework assignments in which students implement and experiment with methods from the lectures. Theoretical labs discuss solutions to assignments made available beforehand.

Prerequisites

Students need a foundation in mathematics comparable to Linear Algebra (B0B01LAG), Calculus (B0B01MA2), Optimization (B0B33OPT), and Probability, Statistics and Theory of Information (B0B01PST).

Beyond mathematics, solid knowledge is expected in basics of graph theory and related algorithms, and in pattern recognition, empirical risk minimization, linear classifiers and support vector machines, as covered in Pattern Recognition and Machine Learning (B4B33RPZ or BE4B33RPZ).

Textbook

I. Goodfellow, Y. Bengio and A. Courville, Deep Learning, MIT Press, 2016.