The Agents’ Selection Methods for a Consensus-Based Investment Strategy in a Multi-agent Financial Decisions Support System

Abstract

Modern financial decision support systems are often based on a multi-agent approach to make advice for investors. However, having a large set of different decisions, collected from agents participating in the process, may entail problems related to data integration and its computational complexity. In this paper, we present some algorithms for selecting agents from a set of all available participants to be included in the eventual decision-making process. All algorithms have been experimentally verified using the a-Trader - a prototype of a multi-agent financial decision support systems on a Forex market.