计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2014年
16期
154-158
,共5页
分形布朗运动%改进Whittle法%R/S%小波分析法%迭代估计法
分形佈朗運動%改進Whittle法%R/S%小波分析法%迭代估計法
분형포랑운동%개진Whittle법%R/S%소파분석법%질대고계법
fractal Brownian motion%improved whittle method%R/S%wavelet analysis%iterative estimation
针对时间序列的Hurst指数的估计方法的问题,目前国内外已经提出了R/S,DAF,绝对值法,周期图法等多种方法。但上述方法都会对Hurst指数的估计值产生易误解和不一致的结果。针对这个问题,通过对R/S分析法,小波分析法,迭代估计算法和Whittle法的描述,进行数值模拟来说明这些方法所得Hurst指数估计值的误差,通过比较能够得出Whittle法是一种具有更高精度和更好稳定性的方法。
針對時間序列的Hurst指數的估計方法的問題,目前國內外已經提齣瞭R/S,DAF,絕對值法,週期圖法等多種方法。但上述方法都會對Hurst指數的估計值產生易誤解和不一緻的結果。針對這箇問題,通過對R/S分析法,小波分析法,迭代估計算法和Whittle法的描述,進行數值模擬來說明這些方法所得Hurst指數估計值的誤差,通過比較能夠得齣Whittle法是一種具有更高精度和更好穩定性的方法。
침대시간서렬적Hurst지수적고계방법적문제,목전국내외이경제출료R/S,DAF,절대치법,주기도법등다충방법。단상술방법도회대Hurst지수적고계치산생역오해화불일치적결과。침대저개문제,통과대R/S분석법,소파분석법,질대고계산법화Whittle법적묘술,진행수치모의래설명저사방법소득Hurst지수고계치적오차,통과비교능구득출Whittle법시일충구유경고정도화경호은정성적방법。
At present, there are many methods of analyzing the Hurst exponent of time series in and abroad, but most of them could lead to misunderstanding and disagreement to the estimation of Hurst exponent. This paper describes R/S analysis, wavelet analysis, iterative algorithm estimation and modified Whittle analysis, by the means of numerical simulation, to compare the error of estimation of Hurst exponent that obtained by the above methods. The comparison is made to test and verify Whittle method has the best accuracy and stability.