Generalized Cauchy Process: Difference Iterative Forecasting Model

The contribution of this article is mainly to develop a new stochastic sequence forecasting model, which also called the difference iterative model based on Generalized Cauchy (GC) process. GC process Long-Range Dependent (LRD) described by two independent parameters: Hurst parameter H and fractal dimension D. Compared with fractional Brownian motion (fBm) linear relationship between D, can more flexibly describe various LRD processes. Before building demonstrates using D properties sequences, respectively. taken as diffusion term establish differential where incremental distribution obtained statistics. parameters are estimated box dimension, rescaled range, maximum likelihood methods. Finally, real wind speed data set used verify performance model.