Work

Complexity-Based Approach to Calibration With Checking Rules

Year: 2011

Type: article

Abstract: We consider the problem of forecasting a sequence of outcomes from an unknown source. The quality of the forecaster is measured by a family of checking rules. We prove upper bounds on the value of the associated game, thus certifying the existence of a calibrated strategy for the forecaster. We show that complexity of the family of checking rules can be captured by the notion of a sequential cover introduced in [19]. Various natural assumptions on the class of checking rules are considered, including finiteness of Vapnik-Chervonenkis and Littlestone's dimensions.

Source: ScholarlyCommons (University of Pennsylvania)

Authors: Dean P. Foster, Alexander Rakhlin, Karthik Sridharan, Ambuj Tewari

Institutions: University of Pennsylvania, Toyota Technological Institute at Chicago, The University of Texas at Austin

Language: English

Topic: Advanced Bandit Algorithms Research

Subfield: Management Science and Operations Research

Field: Decision Sciences

Domain: Social Sciences

Open Access status: green

Funder: National Science Foundation

Awards: CAREER: Statistical and Computational Complexities of Modern Learning Problems, G6671297155

OpenAlex ID: W2182762018

FWCI: 4.161

Cites: 22

Cited by: 14

Related to: 10