Fuzzy Autoregressive Rules: Towards Linguistic Time Series Modeling
介绍模糊规则模型及其与机制转换自回归模型的联系，展示软计算如何帮助实践者更深入理解问题，并用实际波动率序列示例证明其预测能力。
Fuzzy rule–based models, a key element in soft computing (SC), have arisen as an alternative for time series analysis and modeling. One difference with preexisting models is their interpretability in terms of human language. Their interactions with other components have also contributed to a huge development in their identification and estimation procedures. In this article, we present fuzzy rule–based models, their links with some regime-switching autoregressive models, and how the use of soft computing concepts can help the practitioner to solve and gain a deeper insight into a given problem. An example on a realized volatility series is presented to show the forecasting abilities of a fuzzy rule–based model.