Conference paper Open Access

The Megra System - Small Data Music Composition and Live Coding Performance

Reppel, Niklas

This article describes the M ́egra music system, a code-based, stochastic music system that can be used in a live performance context, as well as for longer compositions. M ́egra relies on Probabilistic Finite Automata (PFA) as its fundamental data structure. A case is made for the use of PFAs as a data model that can not only be trained(in the sense of machine learning), but also be interacted with on the basis of predefined operations and, as a side effect, enables one to creatively use the imperfections that occur when using very small datasets to infer musical sequence generators with the help of machine learning methods

Files (1.6 MB)
Name Size
1.6 MB Download
All versions This version
Views 4242
Downloads 3232
Data volume 51.7 MB51.7 MB
Unique views 3939
Unique downloads 2828


Cite as