Learning and Generating Folk Melodies using MPF-Inspired Hierarchical Self-Organising Maps Edwin Hui-Hean Law α and Somnuk Phon-Amnuaisuk α,β α Music Informatics Research Group, β Faculty of Creative Industries, Universiti Tunku Abdul Rahman, Petaling Jaya Campus, Selangor Darul Ehsan, Malaysia. elhh82@gmail.com; somnuk@utar.edu.my Abstract. One of the elements in human music creativity results from certain features in the brain that allows it to make predictions of events based on information learnt from past music experiences. Inspired by the Memory Prediction Framework (MPF) theory, we propose a method to learn and generate new melodies based on the MPF concept. We first show how an MPF-inspired Hierarchical Self Organizing Map (MPF- HSOM) is used to capture these important features of the brain in the perspective of MPF. This MPF-HSOM is then trained with a selection of melodies taken from a corpus of folk melodies. We then show that by using a prediction algorithm, we are able to generate new melodies based on the trained MPF-HSOM of old melodies. The system proposed here is an abstraction of the features of the brain according to MPF. The results indicate that the system is able to learn and to produce novel melodies of reasonable quality. Keywords Hierarchical Self-Organising Map, Learning Folk Melodies, Folk Melody Generation, Memory Predictive Framework 1 Background Advances in neuroscience reveal many interesting facts about our brain and its functions. The columnar organization of the cerebral cortex was first charac- terised by Mountcastle in the 1950s, who, later on in 1978, proposed that the cortical column acted as the unit of computation [11]. Inspired by Mountcas- tle’s work, Hawkins further investigated the concepts of cortical computation units and proposed the Memory Prediction Framework (MPF) [4]. The MPF at- tempts to theorise how the brain functions based on neurological observations. The central idea of the MPF is that the neocortex is a massive memory store. The neocortex processes sensory input which will be used to reinforce previ- ously learned patterns or to form a new learned pattern in neocortical memory. Hawkins suggested four important features of neocortical memory: (i) patterns are stored in sequences ; (ii) patterns are recalled auto-associatively ; (iii) patterns are stored in invariant form, and (iv) patterns are stored in a hierarchy.