On-line Prediction with Kernels and the Complexity Approximation Principle
This work addresses on-line prediction for regression tasks, presenting an incremental extension of earlier linear results to kernel methods.
The paper tackles the problem of on-line regression by applying the Aggregating Algorithm to kernel techniques, resulting in an algorithm that performs nearly as well as any oblivious kernel predictor, with a derived performance estimate.
The paper describes an application of Aggregating Algorithm to the problem of regression. It generalizes earlier results concerned with plain linear regression to kernel techniques and presents an on-line algorithm which performs nearly as well as any oblivious kernel predictor. The paper contains the derivation of an estimate on the performance of this algorithm. The estimate is then used to derive an application of the Complexity Approximation Principle to kernel methods.