An Identity for Kernel Ridge Regression

Abstract

This paper derives an identity connecting the square loss of ridge regression in on-line mode with the loss of the retrospectively best regressor. Some corollaries about the properties of the cumulative loss of on-line ridge regression are also obtained.

35 pages; extended version of ALT 2010 paper (Proceedings of ALT 2010, LNCS 6331, Springer, 2010)

References in corpus (3)

• Competing with Gaussian linear experts
• Prediction with Expert Advice under Discounted Loss
• Sequential optimizing strategy in multi-dimensional bounded forecasting games