-
Notifications
You must be signed in to change notification settings - Fork 2
/
untitled.tex
3 lines (3 loc) · 1.04 KB
/
untitled.tex
1
2
3
\begin{abstract}
This work proposes a decision support system that helps its users choose a stock to buy. In order to arrive to a decision, an architecture of a zero-sum multi-agent system based on fuzzy inference systems and genetic programming was designed. This architecture allows an individual to examine the membership functions of the fuzzy inference systems to better understand the behavior of a financial market. Each of the agents in the multi-agent system uses a fuzzy inference system to decide in what direction they will trade and how much trading force they will use. The multi-agent system performs a regression task on the time series data of a stock, to determine the direction and strength of the next movement in the time series; this process is repeated for every financial market being considered, and the stock with the most predicted strength is recommended to the user to be traded. The method was tested against a stock selection benchmark based on the Dow Jones Index, and the results show that the method is profitable.
\end{abstract}