C4.5: programs for machine learning
C4.5: programs for machine learning
Unifying instance-based and rule-based induction
Machine Learning
Separate-and-Conquer Rule Learning
Artificial Intelligence Review
Clustering Algorithms
Learning Logical Definitions from Relations
Machine Learning
Rule-based machine learning methods for functional prediction
Journal of Artificial Intelligence Research
The Musical Expression Project: A Challenge for Machine Learning and Knowledge Discovery
EMCL '01 Proceedings of the 12th European Conference on Machine Learning
The Musical Expression Project: A Challenge for Machine Learning and Knowledge Discovery
PKDD '01 Proceedings of the 5th European Conference on Principles of Data Mining and Knowledge Discovery
In Search of the Horowitz Factor: Interim Report on a Musical Discovery Project
DS '02 Proceedings of the 5th International Conference on Discovery Science
Automatic performer identification in commercial monophonic Jazz performances
Pattern Recognition Letters
Modeling emotions in violin audio recordings
Proceedings of 3rd international workshop on Machine learning and music
Understanding expressive music performance using genetic algorithms
EC'05 Proceedings of the 3rd European conference on Applications of Evolutionary Computing
| Hi-index | 0.00 |
We present a new rule learning algorithm named PLCG -- a kind of ensemble learning method -- that can find simple, robust partial theories (sets of classification rules) in complex data where neither high coverage nor high precision can be expected. The motivating application problem comes from an interdisciplinary research project that aims at discovering fundamental principles of expressive music performance from large amounts of complex real-world data (measurements of actual performances by concert pianists). It is shown that PLCG succeeds in finding some surprisingly simple and robust performance principles, some of which represent truly novel and musically meaningful discoveries. A more systematic experiment shows that PLCG learns significantly simpler theories than more direct approaches to rule learning, while striking a compromise between coverage and precision.