Exploration of Play Against the Random Past Strategy in Non-Symmetric Sequential Predictions

• Activity Number: 341
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Risk Analysis
• Author(s): Mingfei Li*+
• Companies: Bentley University
• Keywords: play against the random past ; sequential predictions ; Bayes envelope risk ; non-symmetric repeated games

Abstract

In previous studies, play against the random past (PRP) strategy was proposed, and defined. Its asymptotic property Hannan consistency has been studied for symmetric repeated plays, for example, repeated matching binary bits problem, and k-extended Bayes envelope risk problem, for example, two expert-election problem. In this paper, we will begin to explore PRP strategy in non-symmetric sequential predictions. PRP strategy's definition will be generalized to non-symmetric situations. PRP strategy's asymptotic properties will be explored, and studied through the simulations of real Business data.