Output details
11 - Computer Science and Informatics
Queen Mary University of London
Unsupervised Statistical Learning Underpins Computational, Behavioural and Neural Manifestations of Musical Expectation
<22>A novel tripartite approach (behavioural, neurophysiological, and computational) to validating computational cognitive models, and application to our model of melodic expectation. This computational model is now the dominant one in music psychology (15 related papers at ICMPC'12, the relevant biggest conference, http://icmpc-escom2012.web.auth.gr). Also a novel visual cue paradigm for study of time-critical stimuli without artificial pause. The first computational model to predict a quantitative connection between observed behaviour and neural excitation. Led to 2 international keynote presentations (Wiggins: Audio Mostly 2009, http://www.audiomostly.com/index.php?option=com_content&view=article&id=63&Itemid=51; CogMIR 2010 (Cognitive Music Information Retrieval), http://www.cogmir.org/2011-seminar/.) and an EU FP7 grant, "Learning to Create", €2.5M.