Saturday , June 23 2018

An Improved Fuzzy Rule-Based Automated Trading Agent

Universidad Técnica Federico Santa María, Departamento de Informática
Avda. España 1680, Valparaíso, 2390123, Chile

Universidad Adolfo Ibañez, Facultad de Ingeniería y Ciencias
Balmaceda 1625, Viña del Mar, 3132386, Chile

Universidad Adolfo Ibañez, Facultad de Sicología
Balmaceda 1625, Viña del Mar, 3132386, Chile

Universidad Técnica Federico Santa María, Departamento de Informática
Avda. España 1680, Valparaíso, 2390123, Chile

Abstract: In this paper an improved Fuzzy Rule-Based Trading Agent is presented. The proposal consists in adding machine-learning-based methods to improve the overall performance of an automated agent that trades in futures markets. The modified Fuzzy Rule-Based Trading Agent has to decide whether to buy or sell goods, based on the spot and futures time series, gaining a profit from the price speculation. The proposal consists first in changing the membership functions of the fuzzy inference model (Gaussian and Sigmoidal, instead of triangular and trapezoidal). Then using the NFAR (Neuro-Fuzzy Autoregressive) model the relevant lags of the time series are detected, and finally a fuzzy inference system (Self-Organizing Neuro-Fuzzy Inference System) is implemented to aid the decision making process of the agent. Experimental results demonstrate that with the addition of these techniques, the improved agent considerably outperforms the original one.

Keywords: Automated Trading Agents, Fuzzy Rule-based Agents.

>>Full text
Héctor ALLENDE-CID, Enrique CANESSA, Ariel QUEZADA, Héctor ALLENDE, An Improved Fuzzy Rule-Based Automated Trading Agent, Studies in Informatics and Control, ISSN 1220-1766, vol. 20 (2), pp. 135-142, 2011.

1. Introduction

In recent years the existence of many digital mediated tasks that present participants with complex strategic decisions has been more obvious. The agent paradigm has led to decisions that involve a significant interaction between them. The importance of financial trading and electronic commerce can not be ignored, not only for the negotiation and trading involved, but also because of other commercial operations, such as matchmaking, resource finding, advertising, recommendation, contracting and executing transactions [20]. Online markets have set the spotlight in the interest to automate some tasks. There are many reasons why it could be beneficial to keep exploring the automation of financial markets. One advantage is cost saving from automating some functions of the existing non-computational markets. Other could be the exclusion of human behaviour from trading agents. It is known that humans often make emotional decisions, based on biased perceptions of the world. This aspect may be more detrimental when agents need to make fast decisions. Thus, rationalizing the behaviour of agents could be most of the time beneficial from many points of view. Another drawback of human trading agents is that nowadays, the amount of information that is generated is so large, and in such short periods of time, that it is necessary to have agents that can process vast amounts of data and make decisions in seconds, or even in shorter time. This is why we need to have agents with high computational capacity and very fast automated responses. Various works have been proposed in this field, for example [12,14,15,16], where the authors propose and compare automated trading agents that deal with stock trading and supply chain management, among other applications. Additionally, there are many trading agent competitions [1,17], that are designed to promote and encourage high quality research regarding the trading agent problem.

In the last decades fuzzy logic has been an active area of research, many research efforts have been focused on fuzzy modelling and control issues based on the Takagi-Sugeno (TS) fuzzy model [4,6,9,10], which is described by fuzzy IF-THEN rules.

Also the inclusion of this approach to many computational intelligence models has improved them by providing approximate reasoning to models that are generally probabilistic or “crisp” [2, 5,8, 21].

The structure of this work is organized as follows: in the next section, we briefly discuss the fundamentals of Futures Market, present the U-Mart Project and explain the Fuzzy Rule-based trading agent. In section 3 we present our proposal. Then we give some experimental results to show the performance of the proposed agent in section 4. Finally, in the last section, we present the concluding remarks and we delineate some future work.


  1. A. for Trading Agent Research, Online. Available:, September, 2010.
  2. AI-BING, J., P. JIA-HONG, L. SHU-HUAN, S. JIAN-PING, Support Vector Machine for Classification Based on Fuzzy Training Data, Machine Learning and Cybernetics, 2006 International Conference, 13-16 August 2006, pp. 1609-1614.
  3. ALLENDE-CID, H., A. VELOZ, R. SALAS, S. CHABERT, H. ALLENDE, Self-organizing Neuro-fuzzy Inference System, in Congreso Iberoamericano de Reconocimiento de Patrones, 2008, pp. 429-436.
  4. BENREJEB, M., D. SOUDANI, A. SAKLY, P. BORNE, New Discrete TSK Fuzzy Systems Characterization and Stability Domain, International Journal of Computers, Communications & Control, Agora University Editing House – CCC Publications, vol. 1, no. 4, 2006, pp. 9-19.
  5. CHEN, B., L. L. HOBEROCK, A Fuzzy Neural Network Architecture for Fuzzy Control and Classification, Neural Networks, 1996, IEEE International Conference, vol. 2, 3-6 June 1996, pp. 1168-1173.
  6. DIEULOT, J. Y., Fuzzy Control and Estimation Using Model Inversion, Studies in Informatics and Control Journal, ICI Publishing House, vol. 13, no. 3, 2004, pp. 153-160.
  7. GUELY, F., P. SIARRY, Gradient Descent Method for Optimizing Various Fuzzy Rules Bases, in Second IEEE International Conference on Fuzzy Systems, 1993, pp. 1241-1246.
  8. GYEONGYONG, H., P. GADER, Fuzzy SVM for Noisy Data: A Robust Membership Calculation Method, Fuzzy Systems, 2009. FUZZ-IEEE 2009, IEEE International Conference, 20-24 August 2009, pp. 431-436.
  9. HAFAIFA, A., K. LAROUSSI, F. LAAOUAD, Robust Fuzzy Fault Detection and Isolation Approach Applied to the Surge in Centrifugal Compressor Modelling and Control, International Journal of Fuzzy Information and Engineering, Springer, vol. 2 no. 1, 2010, pp. 49-73.
  10. HAFAIFA, A, F. LAAOUAD, K. LAROUSSI, Fuzzy Modelling and Control for Detection and Isolation of Surge in Industrial Centrifugal Compressors, Journal of Automatic Control, University of Belgrade, vol. 19, no. 1, 2009, pp. 19-26.
  11. JANG, J.-S. R., ANFIS: Adaptive-network-based Fuzzy Inference System, IEEE Transactions on Systems, Man and Cybernetics, vol. 23, no. 3, 1993, pp. 665-685.
  12. KEARNS, M., L. ORTIZ, The Penn-Lehman Automated Trading Project, IEEE Intelligent Systems, IEEE Computer Society, 2003.
  13. NAKASHIMA, T., T. ARIYAMA, H. KITANO, H. ISHIBUSHI, A Fuzzy Rule-Based Trading Agent: Analysis and Knowledge Extraction, Computational Intelligence Modelling and Prediction Book, Springer, 2005.
  14. PREIST, C., Economic Agent for Automated Trading, Technical Report, Hewlett Packard Labs, 1998.
  15. SHERSTOV, Y, P. STONE, Three Automated Stock-Trading Agents: A Comparative Study, Agent Mediated Electronic Commerce, Springer, 2005.
  16. SNARSKA, M., J. KRZYCH, Automatic Trading Agent: RMT based Portfolio Theory and Portfolio Selection, Acta Physica Polonica B, vol. 37, no. 11, 2006, pp. 3145-3160.
  17. T.A.Competition, Online. Available:, September, 2010.
  18. U. Systems, Online. Available:, September 2010.
  19. VELOZ, A., H. ALLENDE-CID, R. SALAS, H. ALLENDE, A Flexible Neuro-fuzzy Autoregressive Technique for Non-linear Time Series Forecasting, in Knowledge and Engineering Systems, 2009, pp. 22-29.
  20. WELLMAN, J., Online Marketplaces, in Practical Handbook of Internet Computing, CRC Press, 2004.
  21. ZHENG, P., Q. KEYUN, X. YANG, Dynamic Adaptive Fuzzy Neural-network Identification and Its Application, Systems, Man and Cybernetics, IEEE International Conference, 5-8 October 2003, vol. 5, pp. 4974-4979.