NeuroCOLT

Neural Networks and Computational Learning Theory

 

About NeuroCOLT

Papers Archive

1994 1995
1996 1997
1998 1999
2000 2001

Books

info@neurocolt.org

NeuroCOLT Technical Report NC-TR-00-083

Adaptive and Self-Confident
On-Line Learning Algorithms

Peter Auer
Nicolo' Cesa-Bianchi
Claudio Gentile

Abstract
In this paper, we study the family of quasi-additive learning algorithms applied to on-line linear regression problems. The on-line algorithms in this family are efficient and easy to implement. They work well with various losses and, furthermore, they are very robust to noise, i.e., one can prove that their cumulative loss on any data sequence cannot grow much faster than the loss of the best off-line linear predictor on the same data sequence. However, to achieve these performance bounds, the learning rate must be optimized based on a posteriori information. This information depends on the whole sequence of examples and thus it is not available to any strictly on-line algorithm.

In this paper, we introduce new techniques for adaptively tuning the learning rate as the data sequence is progressively revealed. Our techniques allow us to prove essentially the same performance bounds as if we knew the optimal learning rate in advance. the information needed to optimally tune the learning rate. Moreover, such techniques apply to a wide class of on-line algorithms, including p-norm algorithms for generalized linear regression and Weighted Majority for linear regression with absolute loss.
We emphasize that our adaptive tunings are radically different from previous techniques, such as the so-called doubling trick. Whereas the doubling trick restarts the on-line algorithm several times using a constant learning rate for each run, our methods save information by changing the value of the learning rate very smoothly. In fact, for Weighted Majority over a finite set of experts our analysis provides a better leading constant than the doubling trick.

Download Compressed Postscript