Material Detail

Linear Programming Boosting for Classification of Musical Genre

Linear Programming Boosting for Classification of Musical Genre

This video was recorded at NIPS Workshop on Music, Brain and Cognition, Whistler 2007 . Classification of musical genre from raw audio files is a fairly well researched area of music research, and as such provides a good starting point for testing a new algorithm. The Music Information Retrieval Evaluation eXchange (MIREX) is a yearly competition in a wide range of machine learning applications in music. MIREX 2005 included a genre classification task, the winner of which [1] was an application of the multiclass boosting algorithm AdaBoost.MH [2]. It is believed that Linear Programming Boosting (LPBoost) is a more appropriate algorithm for this application due to the higher degree of sparsity in the solutions [3]. The present study aims to improve on the [1] result by using a similar feature set and the multiclass boosting algorithm LPBoost.MC. References: [1] J. Bergstra, N. Casagrande, D. Erhan, D. Eck, and K. Bal´azs. Aggregate features and ADABOOST for music classification. Machine Learning, 65 (2-3):473–484, 2006. [2] R.E. Schapire and Y. Singer. Improved boosting algorithms using confidence-rated predictions. Machine Learning, 37:297–336, 1999. [3] Ayhan Demiriz, Kristin P. Bennett, and John Shawe-Taylor. Linear programming boosting via column generation. Machine Learning, 46(1–3):225–254, 2002.

Quality

  • User Rating
  • Comments
  • Learning Exercises
  • Bookmark Collections
  • Course ePortfolios
  • Accessibility Info

More about this material

Comments

Log in to participate in the discussions or sign up if you are not already a MERLOT member.