期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2019
卷号：97
期号：3
页码：835-845
出版社：Journal of Theoretical and Applied
摘要：Foreign Exchange market (FOREX) is the global and most liquid market interested in buying, selling and exchanging currencies. The price of these currencies changes rapidly. However, we need to have a good trading strategy to take advantage of these variations. As Forex is a dynamic market, it becomes more difficult to control trading behavior and it becomes very complex to predict the events that can occur. Due to the chaotic, noisy, and non-stationary nature of the data, many algorithmic approaches were adopted in the aim to help traders and make FOREX speculation successful. Algorithmic trading offers the ability to have a strategy in advance. It helps traders to take the final decision. When the decision is so wide and complex, that one algorithm cannot possess all rules to take it. It becomes necessary to call upon several agents, who must work together in pursuing a common objective. These agents co-operate with one another to solve these decision problems. Coordinating between agents in a multi-agent system gives more flexibility and performance to problem solving. Each agent simultaneous results are combined by using a super-agent who helps to make the final decision. In our paper, we propose a theoretical Multi Agent System for stock market Speculation. We use four agents working in one system. The first one is a Metaheuristic Algorithm agent, the second one is based on technical indicators, the third one is a Text Mining agent, and the fourth one is a Fundamental Factor agent. The final decision should be made based on the combination of the four agents results. We think that working with Metaheuristics can improve speculation results compared to the use of classical algorithms. In perspective work, we test our system on a multi-agent systems platform.
关键词：Stock Market; Speculation; Text Mining; Smart Agent; Technical Indicators; Fundamental Factor; Metaheuristic; Particle Swarm Optimization; Multi Agent System
