A Mixed Observability Markov Decision Process Model for Musical Pitch

Abstract: Partially observable Markov decision processes have been widely used to provide models for real-world decision making problems. In this paper, we will provide a method in which a slightly different version of them called Mixed observability Markov decision process, MOMDP, is going to join with our problem. Basically, we aim at offering a behavioural model for interaction of intelligent agents with musical pitch environment and we will show that how MOMDP can shed some light on building up a decision making model for musical pitch conveniently.