Byzantine Music Composition Using Markov Models

Teodor Lucian Anton, Stefan Trausan-Matu
2018 Romanian Conference on Human-Computer Interaction  
This paper presents a project that applies Markov Models improved with similarity metrics to produce computergenerated Byzantine music. An implementation over a dataset of Romanian music in eighth mode, stichiraric style is described. The importance of the similarity mechanism is highlighted, in particular in regard to melodic transition tables. Results of the implementation are presented, which show both strengths and limitations of the model.
dblp:conf/rochi/AntonT18 fatcat:zmiqlyyzjrboneru2xoixaevmu