论文部分内容阅读
本文基于小波-NAR神经网络技术,提出气象要素时间序列预测与天气指数彩虹期权估值的原理与方法,同时采用2000-2014年悉尼日均气温和日降雨量数据,进行气象预测与天气期权估值.结果显示:小波-NAR神经网络因灵活的非线性动态结构较好地反映了气象变化特征,其预测与估值效果优于其他模型;该天气期权价值形成中的非线性特征取决于五种经济效应.科学预测天气和估计天气期权价值,开发天气衍生品,可挖掘天气不确定性的经济价值,弱化其对天气敏感产业的影响.
Based on the wavelet-NAR neural network technology, this paper presents the principle and method of weather element time series forecasting and weather index rainbow option valuation. At the same time, using daily average daily temperature and rainfall data of Sydney from 2000 to 2014, The results show that the nonlinear nonlinear dynamic structure of wavelet-NAR neural network better reflects the characteristics of meteorological changes, and its prediction and evaluation effect is superior to other models. The nonlinear characteristics of the weather option value formation depend on five Economic effects: Scientifically forecasting the weather and estimating the value of weather options, developing weather derivatives can tap the economic value of weather uncertainty and weaken its impact on weather-sensitive industries.