Instructor: Rafael Ramirez
Credits: 5 ECTS

Machine learning has proved to provide efficient solutions to many music-related problems both of academic and commercial interest. This course of the Master in Sound and Music Computing focuses on machine learning, data mining and pattern recognition techniques and their application to audio and music technology. The expected outcome of the course is a deep understanding of the main machine learning techniques and their relevance in audio and music applications. The course promotes and provides opportunities to present and discuss ongoing and state-of-the-art research in the area.

The course is offered in 20 weeks, with 30 hours of lectures and seminars. The evaluation of the students is based on oral presentations (60%), assignments (30%), and student participation (10%).

All the materials prepared for the class are available in: https://wikis.dtic.upf.edu/wikis/jclub/

Topics covered

  • Introduction to Machine Learning
  • Retrieval Evaluation
  • Statistics and Statistical Inference
  • Hidden Markov Models
  • Support Vector Machines
  • Decision/Regression Trees
  • Instance-based Learning: k Nearest Neighbor and Case-based Reasoning
  • Artificial Neural Networks
  • Genetic Algorithms/Programming
  • Inductive Logic Programming
  • Clustering
  • Ensemble Methods
  • Feature Selection
  • Principal Component Analysis
  • Dynamic Time Warping
  • Applications to Music Technology

Materials

Software for the course: Matlab, R, Weka