Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Online Algorithm for Aggregating Experts' Predictions with Unbounded Quadratic Loss

Alexander Korotin, Vladimir V'yugin, Evgeny Burnaev

TL;DR

This work considers the problem of online aggregation of expert predictions with the quadratic loss function and proposes an algorithm for aggregating expert predictions which does not require a prior knowledge of the upper bound on the losses.

Abstract

We consider the problem of online aggregation of expert predictions with the quadratic loss function. We propose an algorithm for aggregating expert predictions which does not require a prior knowledge of the upper bound on the losses. The algorithm is based on the exponential reweighing of expert losses.

Online Algorithm for Aggregating Experts' Predictions with Unbounded Quadratic Loss

TL;DR

This work considers the problem of online aggregation of expert predictions with the quadratic loss function and proposes an algorithm for aggregating expert predictions which does not require a prior knowledge of the upper bound on the losses.

Abstract

We consider the problem of online aggregation of expert predictions with the quadratic loss function. We propose an algorithm for aggregating expert predictions which does not require a prior knowledge of the upper bound on the losses. The algorithm is based on the exponential reweighing of expert losses.
Paper Structure (1 theorem, 1 algorithm)

This paper contains 1 theorem, 1 algorithm.

Key Result

Theorem 1

The regret of Algorithm algorithm-unbounded satisfies $R_{T}\leq O(\max\limits_{t,n}l_{t}^{n}\cdot (\ln N +1))$.

Theorems & Definitions (2)

  • Theorem 1
  • proof