Patterns for Prediction task @ MIREX2019

If this is the type of topic that you dread, just tell me so that I may be able to perform some better post prediction in the phuture.

Calling on all pre-cogs:

With apologies for cross-posting…

Call for participation: Patterns for Prediction task @ MIREX2019

https://www.music-ir.org/mirex/wiki/2019:Patterns_for_Prediction

We are searching for researchers whose computational models and/or algorithms have the ability to predict the next musical events (continuation), based on given, foregoing events (prime). We are also interested in models that may not make explicit predictions, but can estimate the likelihood of several alternative continuations.

One facet of human nature comprises the tendency to form predictions about what will happen in the future. Music provides an excellent setting for the study of prediction, and we hope that this task will attract interest from fields such as psychology, neuroscience, music theory, music informatics, and machine learning.

Why “patterns” in “Patterns for Prediction”? This task emerged from an existing Pattern Discovery task, which ran 2013-17. The last five years have seen an increasing interest in discovering or generating patterned data, leveraging methods beyond typical (e.g., Markovian) limits. E.g., how might exact and inexact repetition, occurring over the short, medium, and long term in pieces of music, interact with expectations in order to form a basis for successful prediction?

MIREX stands for the Music Information Retrieval Evaluation eXchange. Since 2005, it has provided a forum for researchers to (1) train algorithms to perform specific, music-technological tasks on publicly available datasets, (2) submit algorithms that are run and evaluated by MIREX organizers on private datasets, (3) compare their work with one another and shed light on research questions informed by and informing diverse fields intersecting with music informatics.

The deadline for submitting to this task is Monday September 9th, 2019. If you are interested in participating in this task but do not think you will have time until the 2020 iteration, please let us know so we can keep you in mind for next year.

For more details, please refer to the MIREX page: https://www.music-ir.org/mirex/wiki/2019:Patterns_for_Prediction

Your task captains are Iris Yuping Ren (yuping.ren.iris), Berit Janssen (berit.janssen), James Owers (james.f.owers), and Tom Collins (tomthecollins, all at gmail.com). Feel free to copy in all four of us if you have questions/comments.

Thanks for reading!