期刊名称：Analele Universităţii Constantin Brâncuşi din Târgu Jiu : Seria Economie
印刷版ISSN：1844-7007
出版年度：2012
卷号：4
出版社：Academica Brâncuşi
摘要：Because of sensitive dependence on initial conditions (SDIC), characteristic to chaotic systems, the predictionof such system can be made with an accepted accuracy only for relatively small number of steps ahead. Using artificialtechniques like neural networks and support vector machine to predict chaotic dynamics present advantages overtraditional methods and usually they offers better results. In this paper, I highlight some of these advantages
关键词：Chaos Theory; Artificial Intelligence; Time Series; Exchange rate
