Course Introduction

      This course deals with the algorithmic annotation, understanding, recognition, and categorization of music in symbolic (score-like) form, as well as methodology to accomplish these tasks. Particular applications will be taken from key finding, harmonic analysis, note spelling, rhythm recognition, meter induction, piano fingering, melody morphing, and various classification problems such as genre or composer identification. The methodology we will employ will be probabilistic and will include ideas from machine learning such as optimal classifiers, hidden Markov models, and possibly some Bayesian networks. Students will have computing assignments in R (similar to Matlab) and will occasionally present papers. We will employ the "learn by doing" method throughout the course. The course will be supported by readings of various papers and tutorials. [Course Outline]

Instructor C. Raphael (
Classes Tue/Thu, 1:00-2:15pm, M 340
Office Informatics 315, T.(812)-856-1849
Office Hours Mon/Wed, 4:00-5:00pm by appointment
AI Yupeng Gu (
Office Hours

Thu 2:30-4:00pm @ Informatics 313 OR by appointment

UPDATE(11/13) 6th Homework Assignment Due Nov.25th
UPDATE(11/06) functional harmonic analysis with HMM examples
UPDATE(10/28) 5th Homework Assignment Due Nov.6th
UPDATE(10/23) How to play audio in R
UPDATE(09/09) Course Notes