What's New?
Gunnar Rätsch, Dr.
Publications
Using Support Vector Machines for Time Series Prediction
Document Actions
G Rätsch, T Onoda, and K Müller (1998)
Soft Margins for AdaBoost
Royal Holloway College, NeuroCOLT, Technical Report(NC-TR-1998-021), University of London.
This paper shows that although AdaBoost rarely overfits in the low noise regime it clearly does so for higher noise levels. We propose several regularization methods and generalizations of the original AdaBoost algorithm to achieve a soft margin – a concept known from Support Vector learning. Extensive simulations demonstrate that the proposed regularized AdaBoost-type algorithms are useful and competitive for noisy data.
to appear in Machine Learning
The group will join the Sloan-Kettering Institute in New York City in January 2012
