LFU:online

The aim is to understand the theoretical concepts and models introduced in the course, with applications to the modelling of complex adaptive systems.

The target audience consists of interested and motivated students from advanced Bachelor, Master and PhD studies.

This course is concerned with systems which are equipped with memory and which interact with their environment through a percept-action loop - referred to as complex adaptive systems (examples include biological life and artificial intelligence systems). This is an interdisciplinary course on the theoretical modelling of informational aspects of such complex adaptive systems.

The first part of this course focuses on a selection of information-theoretic tools and models including 

• an axiomatic approach to surprisal, entropy, and mutual information
• stochastic processes and channels
• communication capacity of channels
• Bayesian inference, Bayesian networks, particle filters, variational inference
• percept-action loops and hidden Markov models
• active variational inference, variational free energy principle
• minimal sufficient statistics, epsilon machines

The second part of this course covers topics from quantum information theory including:

• mathematical properties of the set of quantum states and measurements
• purification, Stinespring dilation, Kraus representation, Choi isomorphism
• instruments, quantum combs, link product
• communication capacity of quantum channels
• quantum Markov chains

Lectures with exercises.

Will be discussed in the first lecture.

Completion of an undergraduate quantum physics course is recommended for the second half of this course.

The course will take place Wed 17:15-18:45 and Thu 08:30-10:00. Every fourth time slot will be replaced with an optional exercise seminar.