AUTHORS: S. Sakha, S. Boonthiem, W. Klongdee
Download as PDF
ABSTRACT: In this paper, the new forecasting algorithm for predicting gold prices using concept of fuzzy logic with membership function model as probability density function of normal distribution is proposed and compared to the triangular membership function model and the trapezoidal membership function model. The daily gold prices from January 2015 to May 2018 collected from the London Bullion Market Association are used in this study. The prices are transformed to the rate of return and then the Train : Test percent ratio is used to obtain the training data and the testing data, respectively, and consists of 99:01, 95:05, 90:10, 80:20, 70:30, 60:40, 50:50, 40:50, 30:70, 20:80 and 10:90. The root mean squared error and the mean absolute error are applied to measure the performance of the proposed algorithm and to compare three presented models. The experimental results can conclude that the proposed algorithm can be practical to the gold price forecasting; furthermore, the proposed algorithm with the membership function model as probability density function of normal distribution can better improve itself than other models when it has more training data as well as it can demonstrate better performance than others
KEYWORDS: Fuzzy logic, gold price forecasting, probability density function of normal distribution, trapezoidal membership function, triangular membership function
REFERENCES:
[1] C. Pierdzioch, M. Risse, and S. Rohloff. “On
the efficiency of the gold market: Results of a
real-time forecasting approach,” Int. Review of
Fin. Analysis, vol. 32, pp. 95–108. Jan. 2014.
[2] R. Gupta, S. Hammoudeh, W. J. Kim, and B.
D. Simo-Kengne. “Forecasting China‟s foreign
exchange reserves using dynamic model
averaging: The role of macroeconomic
fundamentals, financial stress and economic
uncertainty,” N. Am. Journ. of Econ. and Fin.,
vol. 28, pp. 170–189. Feb. 2014.
[3] L. A. Zadeh. “Fuzzy Sets,” Information and
Control, vol. 8, pp. 338 – 353. 1965.
[4] H. A. Sturges. “The Choice of a Class
Interval,” Journ. of the Am. Statis., vol. 21, No.
153, pp. 65–66. Mar. 1926.
[5] R. J. Hyndman. (1995, January). The problem
with Sturges‟ rule for constructing histograms.
Available:
https://robjhyndman.com/publications/sturges/
[6] C. Altavilla, and P. de Grauwe. “Nonlinearities
in the relation between the exchange
rate and its fundamentals,” Int.l Journ. of Fin.
and Econ., vol. 15, No. 1. pp. 1–12. 2010.
[7] C. Beneki, and E. S. Silva. “Analysing and
Forecasting European Union Energy Data,” Int.
Journ. of Energy and Statis., pp. 1–16. Jun.
2013.
WSEAS Transactions on Business and Economics, ISSN / E-ISSN: 1109-9526 / 2224-2899, Volume 15, 2018, Art. #33, pp. 340-347
Copyright Β© 2018 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0