Through theoretical analyzing the classical Farid high order statistic linear prediction model based on discrete Haar wavelet,a R-Farid linear prediction model is proposed by revising its inconsistent relativity area,and then it is defined as Farid high order statistic model,a R-Farid high order statistic model is designed.

Method of embedding dimension estimationbased on high-order statistics;

Image recognition method based on high-order statistics;

Application of high-order statistics to identify weak seismic signal and "magnetic bright spot".;

Coherency algorithm based on higher-order statistics and sedimentary evolvement analysis of sandstone bodies;

Applications of higher-order statistics to the study of pressure fluctuation signal of gas-solid fluidized bed;

Modeling and simulation of wind speed on wind farm by applying higher-order statistics;

Identification and source separation algorithm for MIMO system based on higher order statistics;

A comparative study of blind speech source separation algorithms based on higher order statistics and second order statistics;

A Weighted Image Hybrid Filter Based on Higher Order Statistics;

A texture image recognizing method based on high order statistics;

Using the property that high order statistics are not sensitive to Gaussian noise to processing gaussian noise,and can mostly save the image details which greatly dwindling noise.

And how to use these high order statistics in determination of high order correlation and thus in detection .

Higher-order Statistics (HOS) is the primary analysis tool in analyzing non-Gaussian signal and nonlinear signal.

Higher-order Statistics (HOS) is the primary analysis tools which can be used to analyze noneasual system, nonminimum phase system and nongaussian signal.

A novel class of analysis tool in signal processing-Higher-Order Statistics (HOS) is presented in this thesis.