最大熵分位值估计及其对偶型优化解法

吴福仙; 温卫东

doi:10.13224/j.cnki.jasp.2017.02.020

最大熵分位值估计及其对偶型优化解法

doi: 10.13224/j.cnki.jasp.2017.02.020

南京航空航天大学能源与动力学院,南京 210016

计量
- 文章访问数: 838
- HTML浏览量: 3
- PDF量: 428
- 被引次数: 0
出版历程
- 收稿日期: 2015-05-22
- 刊出日期: 2017-02-28

Maximum entropy quantile estimation and dual optimization method

摘要

摘要: 针对经典最大熵分位值估计中拉格朗日系数计算目前存在高度非线性、计算结果精度不高或有时难以收敛等问题,提出了一种对偶型逐次寻优的方法.基于拉格朗日对偶法,推导建立了含有拉格朗日系数优化函数的对偶表达式；在此基础上,基于样本的概率权重矩约束,提出了逐次寻优算法.针对几种常见的概率分布类型和一种较为复杂的概率分布类型,采用对偶型最大熵方法和经典最大熵方法对其概率累积函数和分位值进行计算对比分析表明:对偶型最大熵分位值估计不仅具有非线性程度低、形式简单,而且对偶型逐次寻优的方法具有比较高的计算精度,优化迭代的收敛性好等特点.
- 分位值估计 /
- 对偶型 /
- 最大熵 /
- 逐次优化 /
- 概率权重矩
Abstract: Based on high nonlinearity, low computational accuracy, or hard convergence of Lagrangian calculation in the classic maximum entropy quantile function, a new method called dualsequential updating method was proposed. Dual expression of optimization function was established on the basis of Lagrangian dual approach. Further sequential updating method was proposed based on the constraint by the probability weighted moments. After analyzing several common distributions and a complex distribution by dual maximum entropy quantile function and classic maximum entropy quantile function, result showed that,dual maximum entropy quantile function had advantage of low nonlinearity and simple form, and dualsequential updating method had good computational accuracy and optimization convergence.
- quantile estimation /
- dual /
- maximum entropy /
- sequential updating /
- probability weighted moments

HTML

参考文献(18)

[1]	Pandey M D.Direct estimation of quantile functions using the maximum entropy principle［J］.Structural Safety,2000,22(1):61-79.
[2]	袁修开,吕震宙,岳珠峰.小样本下分位数函数的Bootstrap置信区间估计［J］.航空学报,2012,33(10):1842-1849. YUAN Kaixiu,LU Zhenzhou,ZHU Yuefeng.Bootstrap confidence interval of quantile function estimation for small samples［J］.Acta Aeronautica et Astronautica Sinica,2012,33(10):1842-1849.(in Chinese)
[3]	Wu X.Calculation of maximum entropy densities with application to income distribution［J］.Journal of Econometrics,2003,115(2):347-354.
[4]	Shi X,Teixeira A P,Zhang J,et al.Structural reliability analysis based on probabilistic response modelling using the maximum entropy method［J］.Engineering Structures,2014,70(9):106-116.
[5]	Hehua Z,Yulong Z,Xiaojun L.Estimation of the fracture diameter distributions using the maximum entropy principle［J］.International Journal of Rock Mechanics and Mining Sciences,2014,72:127-137.
[6]	Li C,Wang W,Wang S.Maximum entropy monte carlo method for the evaluation of dam overtopping probability［J］.Disaster Advances.2012,5(4):1143-1147.
[7]	邓建,古德生,李夕兵.确定可靠性分析Weibull分布参数的概率加权矩法［J］.计算力学学报,2004,21(5):609-613. DENG Jian,GU Desheng,LI Xibing.Parameters and quantile estimation for fatigue life distribution using probability weighted moments［J］.Chinese Journal of Computational Mechanics,2004,21(5):609-613.(in Chinese)
[8]	Deng J,Pandey M D.Estimation of minimum crossentropy quantile function using fractional probability weighted moments［J］.Probabilistic Engineering Mechanics,2009,24(1):43-50.
[9]	Caeiro F,Ivette G M,Vandewalle B.Semiparametric probabilityweighted moments estimation revisited［J］.Methodology and Computing in Applied Probability,2014,16(1):1-29.
[10]	Li X,Zuo Y,Zhuang X,et al.Estimation of fracture trace length distributions using probability weighted moments and Lmoments［J］.Engineering Geology,2014,168(2):69-85.
[11]	ZHANG Ming,BAI Shaoquang.Application of probabilityweighted mixed moment method to parameters estimation for threeparameter probability distribution［J］.Water Resources and Power,2013,31(8):16-18.
[12]	ZHANG Ming,BAI Shaoguang.Probabilityweighted mixed moments method and its application to estimation parameters of pearson type:Ⅲ distribution［J］.Water Resources and Power,2012,30(9):20-29.
[13]	Deng J,Pandey M D,Xie W C.Maximum entropy principle and partial probability weighted moments［R］.Washington D C:American Institute of Physics,2012.
[14]	Deng J,Pandey M D.Derivation of sample oriented quantile function using maximum entropy and selfdetermined probability weighted moments［J］.Environmetrics,2010,21(2):113-132.
[15]	Deng J,Pandey M D.Using partial probability weighted moments and partial maximum entropy to estimate quantiles from censored samples［J］.Probabilistic Engineering Mechanics,2009,24(3):407-417.
[16]	Yu L J,Zhang X N.Maximum entropy models for mode split forecasting［R］.New York:International Conference on Service Operations,Logistics and Informatics,2009.
[17]	Yu L J.A new calculation algorithm for quantile function with application to headway distribution［R］.New York:International Conference on Service Operations,Logistics and Informatics,2008.
[18]	Deng J,Pandey M D.Estimation of the maximum entropy quantile function using fractional probability weighted moments［J］.Structural Safety,2008,30(4):307-319.