HFIMV9052 – Music and Machine Learning

Schedule, syllabus and examination date

Course content

The aim of the course is to develop knowledge of and practical experience with machine learning algorithms applied to music analysis, music information retrieval, interactive music systems, and generative music.

Learning outcome

Having completed the course, the student will:

  • know about various techniques for supervised, unsupervised and reinforcement machine learning.

  • know different feature extraction methods for sound, music and sensor data.

  • be familiar with generic and audio-specific techniques for data mining in music databases.

  • be able to use machine learning techniques for pragmatic and creative purposes in the broad context of music.

  • be able to carry out content-based search in audio collections using music information retrieval techniques.

  • be able to use techniques for action and gesture recognition in interactive music systems.

  • be able to critically reflect on the use of machine learning techniques in applications within and outside the field of music.

  • have an overview of the research frontier in the field.

Admission to the course

The course is open to Ph.D. candidates who are interested in one or more topics in this course. Interested candidates must apply each semester. It is assumed that the applicants have admission to a Ph.D. program at UiO or at another educational institution. The department`s own Ph.D. candidates are given first priority.

All Ph.D. candidates register for the course by notifying the department`s Research Consultant.

External applicants must submit the following documents before teaching begins.

  • confirmation that you have been admitted to a Ph.D. program

  • a brief statement of motivation, containing information about your Ph.D. program, your dissertation and why this topic is relevant to you.

Documentation and questions can be sent to the Research Consultant at IMV.

It is recommended that the student are familiar with Python programming with packages for scientific computing, and have some knowledge in sound and music computing.

Overlapping courses

Teaching

The course is taught using a flipped classroom model and blended learning methods, and includes:

  • Video lectures, readings and assignments in preparation for the workshops.
  • 10 workshops of 4 hours with compulsory attendance.

In order to qualify for the exam, students must have completed all assignments and have obtain at least 80% attendance.

Examination

  • Portfolio (semester project)??

    • The project consists in the design and implementation of a music-related machine learning system. This may include systems for music classification, sound recognition, music recommendation, music database mining, algorithmic music composition, sound processing or generation. Projects are decided by the students and must be approved by the course responsible. Produced project material must be submitted with documentation to replicate the results. Projects are summarized in an academic paper describing design, implementation and evaluation of the developed system, with a particular focus on related works, selected machine learning techniques, dataset, evaluation strategy, and consideration of associated musical aspects. The paper should also include a thorough literature review as well as critical thoughts and reflections.

    • The body of the paper should be approximately 4000 words.

The semester paper must be submitted to the course instructor by the given deadline.

Language of examination

The examination text is given in English, and you submit your response in English.

Grading scale

Grades are awarded on a pass/fail scale. Read more about the grading system.

More about examinations at UiO

You will find further guides and resources at the web page on examinations at UiO.

Last updated from FS (Common Student System) Nov. 5, 2024 10:55:06 AM

Facts about this course

Level
PhD
Credits
5
Teaching
Spring
Examination
Spring
Teaching language
English

Contact

Department of Musicology