Sunday, August 18, 2013

Symbolic Regression for Data Modeling

Overview
Symbolic Regression allows domain experts to create, add modify rules or policies extracted from data. The most commonly used algorithms used in Symbolic Regression are:
- Genetic Algorithms
- Learning Classifiers Systems

Symbolic regression is used in many application ranging from network performance optimization, predicting failure (MTBF), streaming data to detecting security breaches.


Presentation
The following presentation describes the main components, benefits and drawbacks of symbolic regression.



References
Genetic Programming: on the Programming of Computers by Means of Natural Selection - J. Koza - MIT Press 1992
Reinforcement Learning: An introduction (Adaptive Computation and Machine learning) - R. Sutton, A. Barto - MIT Press 1998

No comments:

Post a Comment