This class is usually taught in the winter term. The class is taught in English.

Information for the current semester (if available):

This course reviews the fundamentals of musical informatics with an emphasis on computational modeling of human cognition, and introduces some of the standard algorithms for symbolic music processing. This course is aimed at students that are interested in the intersection of music, mathematics, artificial intelligence and cognitive sciences. 

In this course we'll ask questions like: 

• How can we represent music symbolically (e.g., musical scores)?
• What is the relation between musical symbols and our perception of music?
• How can we analyze the structure of the music with computational methods?
• How can we find rules that generate a new piece of music from sequences of discrete symbols?
• How can two musical sequences be aligned?
• How can we compare how similar are musical sequences?
• What is the probability of occurrence of a musical event given its musical context?

For this course we will introduce some of the standard computational methods used in computational music analysis (e.g., dynamic time warping, Markov models, formal grammars, genetic algorithms, artificial neural networks etc.).

Note that this course focuses on symbolic music processing (e.g., MIDI, MusicXML) rather than audio.

Prerequisites

The course will be self-contained, but it would be useful to have some basic knowledge of probability theory and linear algebra.

Furthermore, it would be useful to have some programming experience (in particular Python) and to be able to read music (at least a little). That being said, knowledge of music theory is not required ;) 

Prior knowledge in these areas may be helpful but is not a requirement.

Course Outline

The course will consist of three parts

1. A general overview of music theory and music cognition. We will start with quick overview of the basic elements of (Western) music theory (e.g., intervals, chords, tonality,...), as well as a review of the fundamentals of music cognition.
2. Computational music analysis. In this part we will present some of the standard problems in computational music analysis (e.g., key identification, music alignment etc.).
3. Music generation. The final part of the course will focus on the use of computational methods to generate new music.

Grading

The coursework will consist of two practical projects, the first one focusing on an aspect of computational music analysis and the second one on generation of music, each of which will be 50% of the total grade. 

The topics for the projects will be presented during the first session, and students will be able to choose their preferred topics.

The results of their projects have to be submitted as a blog post/presentation using self-contained Jupyter Notebooks.

Teaching Materials

A PDF version of the slides will be made available via Moodle.

Bibliography

M. Müller, Fundamentals of Music Processing, 2nd edition, Springer, 2021 

D. Meredith (ed.), Computational Music Analysis, Springer, 2016

D Deutsch (ed.), The Psychology of Music, 3rd edition, Academic Press, 2013

F. Lerhdal and R. Jackendoff, Generative Theory of Tonal Music, MIT Press, 1983

D. Huron, Sweet Anticipation: Music and the Psychology of Expectation,  MIT Press, 2006

D. Temperley, Music and Probability, MIT Press, 2010

G. Nierhaus, Algoritmic Composition: Paradigms of Automated Music Generation, Springer, 2009