Financial time series forecasting based on wavelet and multi-dimensional Taylor network dynamics model
作者： Zhou B
DOI:
关键词: Identification (information) 、 Conjugate gradient method 、 Wavelet 、 Series (mathematics) 、 Artificial intelligence 、 Algorithm 、 Network dynamics 、 Network model 、 Computer science 、 Multi dimensional 、 SIGNAL (programming language)
摘要: Presented in this paper is a new approach to establishment of the dynamics model multidimensional Taylor network and its parameter identification,whereby method based on wavelet for financial time-series forecasting proposed.The time series are decomposed into sub-series low frequency signal several high signals via Mallat algorithm,for each which multi-dimensional established.Model parameters trained by conjugate gradient method,and then models used forecasting.All results superimposed obtain predicted value original series.As verified our experiments,the proposed works well ensuring accuracy forecasting.
cnki.com.cn参考文章(0)