针对一种具浴盆型失效率曲线的寿命分布,本文在记录值数据场合研究了模型参数的置信集估计问题。首先,基于记录值样本构造枢轴量,进而分别建立模型参数的精确置信区间和置信域。其次,为补充起见,进一步利用大样本原理构造了模型参数的...针对一种具浴盆型失效率曲线的寿命分布,本文在记录值数据场合研究了模型参数的置信集估计问题。首先,基于记录值样本构造枢轴量,进而分别建立模型参数的精确置信区间和置信域。其次,为补充起见,进一步利用大样本原理构造了模型参数的渐近置信区间估计。最后,通过数值模拟和算例分析比较了不同置信集结果的优良性。This paper investigates the confidence set estimation problem of model parameters for a lifetime distribution with a bathtub-shaped failure rate curve in the case of recorded data. Firstly, based on the recorded value samples, pivotal quantities are constructed, and then exact confidence intervals and confidence regions for the model parameters are established separately. Secondly, for the sake of supplementation, the asymptotic confidence interval estimation of the model parameters was further constructed using the large sample principle. Finally, the superiority of different confidence set results was compared through numerical simulation and case analysis.展开更多
样本有限的表格型数据缺乏不变性结构和足够样本,使得传统数据增强方法和生成式数据增强方法难以获得符合原始数据分布且具有多样性的数据.为此,文中依据表格型数据的特点和邻域风险最小化原则,提出基于邻域分布的去噪扩散概率模型(Vici...样本有限的表格型数据缺乏不变性结构和足够样本,使得传统数据增强方法和生成式数据增强方法难以获得符合原始数据分布且具有多样性的数据.为此,文中依据表格型数据的特点和邻域风险最小化原则,提出基于邻域分布的去噪扩散概率模型(Vicinal Distribution Based Denoising Diffusion Probabilistic Model,VD-DDPM)及相应算法.首先,分析样本有限表格型数据的特征,通过先验知识选择弱相关特征,并构建样本的邻域分布.然后,利用邻域分布采样数据构建VD-DDPM模型,并使用VD-DDPM数据生成算法生成符合原始数据分布且具有多样性的数据集.在多个数据集上针对数据生成质量、下游模型性能等进行实验,验证VD-DDPM的有效性.展开更多
文摘针对一种具浴盆型失效率曲线的寿命分布,本文在记录值数据场合研究了模型参数的置信集估计问题。首先,基于记录值样本构造枢轴量,进而分别建立模型参数的精确置信区间和置信域。其次,为补充起见,进一步利用大样本原理构造了模型参数的渐近置信区间估计。最后,通过数值模拟和算例分析比较了不同置信集结果的优良性。This paper investigates the confidence set estimation problem of model parameters for a lifetime distribution with a bathtub-shaped failure rate curve in the case of recorded data. Firstly, based on the recorded value samples, pivotal quantities are constructed, and then exact confidence intervals and confidence regions for the model parameters are established separately. Secondly, for the sake of supplementation, the asymptotic confidence interval estimation of the model parameters was further constructed using the large sample principle. Finally, the superiority of different confidence set results was compared through numerical simulation and case analysis.
文摘样本有限的表格型数据缺乏不变性结构和足够样本,使得传统数据增强方法和生成式数据增强方法难以获得符合原始数据分布且具有多样性的数据.为此,文中依据表格型数据的特点和邻域风险最小化原则,提出基于邻域分布的去噪扩散概率模型(Vicinal Distribution Based Denoising Diffusion Probabilistic Model,VD-DDPM)及相应算法.首先,分析样本有限表格型数据的特征,通过先验知识选择弱相关特征,并构建样本的邻域分布.然后,利用邻域分布采样数据构建VD-DDPM模型,并使用VD-DDPM数据生成算法生成符合原始数据分布且具有多样性的数据集.在多个数据集上针对数据生成质量、下游模型性能等进行实验,验证VD-DDPM的有效性.
