宽象限相依下密度估计的一致强相合性OA
Uniform Strong Consistency of Density Estimation for Widely Orthant Dependent Samples
频率多边形插值估计是一种重要的概率密度估计方法,其收敛速度与核密度估计相当,但要优于直方图估计,在计算复杂度和边界处理上更具优势.特别是在处理二元数据集时,频率多边形插值估计不仅计算简便,还能够更便捷地确定确切的等概率轮廓,因而相比核密度估计具有更加突出的研究价值.文章聚焦于相依数据背景下频率多边形估计的理论性质,基于宽象限相依序列(widely orthant dependent,WOD)的Bernstein型指数不等式和截尾技术,系统探讨了未知密度函数的频率多边形插值估计及其在可靠性分析中至关重要的失效率函数估计.该研究在适当条件下证明了估计量的一致强相合性,推广了已有文献的相关结论,深化了相依样本下非参数密度估计的理论体系.
The interpolated frequency polygon estimator is an important method for probability density estimation.Its convergence rate is comparable to that of kernel density estimation but superior to that of histogram estimation,offering dis-tinct advantages in computational complexity and boundary treatment.This advantage is particularly evident in bivariate set-tings,where it facilitates the construction of exact equiprobable contours.Focusing on the theoretical properties of frequency polygon estimators in the context of dependent data,this paper establishes the theoretical properties of frequency polygon esti-mators under widely orthant dependent(WOD)sequences.Utilizing Bernstein-type exponential inequalities and truncation techniques,we derive the uniform strong consistency for both the density estimator and the associated failure rate function,a key metric in reliability analysis.Our results extend existing literature and reinforce the theoretical foundation of nonparamet-ric estimation for dependent samples.
李会葆;潮学琳;阮景
安庆师范大学 数学与统计学院,安徽 安庆 246133安庆师范大学 数学与统计学院,安徽 安庆 246133安庆师范大学 数学与统计学院,安徽 安庆 246133
数理科学
WOD频率多边形失效率函数一致强相合性密度估计
WODfrequency polygonfailure rate functionuniformly strong consistencydensity estimation
《安庆师范大学学报(自然科学版)》 2026 (2)
28-32,5
大别山区域复杂生态系统建模、仿真与控制重点实验室项目(SLXY24A001),安徽省高等学校省级质量工程项目(2025zytsct029)和安庆师范大学校级质量工程项目(2024aqnujyxm07)
评论