魏正元, 彭天奎, 周晓娅. Alpha偏幂正态分布及其应用[J]. 应用概率统计, 2022, 38(5): 647-658. DOI: 10.3969/j.issn.1001-4268.2022.05.002
引用本文: 魏正元, 彭天奎, 周晓娅. Alpha偏幂正态分布及其应用[J]. 应用概率统计, 2022, 38(5): 647-658. DOI: 10.3969/j.issn.1001-4268.2022.05.002
WEI Zhengyuan, PENG Tiankui, ZHOU Xiaoya. Alpha Skew Power Normal Distribution and Its Application[J]. Chinese Journal of Applied Probability and Statistics, 2022, 38(5): 647-658. DOI: 10.3969/j.issn.1001-4268.2022.05.002
Citation: WEI Zhengyuan, PENG Tiankui, ZHOU Xiaoya. Alpha Skew Power Normal Distribution and Its Application[J]. Chinese Journal of Applied Probability and Statistics, 2022, 38(5): 647-658. DOI: 10.3969/j.issn.1001-4268.2022.05.002

Alpha偏幂正态分布及其应用

Alpha Skew Power Normal Distribution and Its Application

  • 摘要: 提出了一类新的alpha偏幂正态分布,使得alpha偏正态分布和幂正态分布是其特例. 给出了该新分布的矩、偏度、峰度、Fisher信息阵以及参数的最大似然估计量、随机数的产生算法等.大量的随机模拟试验表明, 参数最大似然估计量具有较好的渐近统计性质,且对一组真实数据集的分布拟合结果显示, alpha偏幂正态分布比正态分布、偏正态分布、幂正态分布和alpha偏正态分布的拟合效果更好.

     

    Abstract: A new class of alpha skew power normal distribution is proposed, such that alpha skew normal distribution and power normal distribution are special cases. Its moment, skewness, kurtosis, Fisher information matrix, maximum likelihood estimator and the generation algorithm of random number are given. Extensive simulation experiments show that the maximum likelihood estimator of parameters has good asymptotic statistical properties, the distribution fitting results of a group of real data sets show that alpha skew power normal distribution has better fitting effect than normal distribution, skew normal distribution, power normal distribution and alpha skew normal distribution.

     

/

返回文章
返回