Journal of Statistical Planning and Inference ( IF 0.8 ) Pub Date : 2021-06-22 , DOI: 10.1016/j.jspi.2021.06.001 Péter Kevei , Lillian Oluoch , László Viharos
Denote , where , is a sequence of integers such that and , and are the order statistics of iid random variables with regularly varying upper tail of index . The estimator is an extension of the Hill estimator. We investigate the asymptotic properties of and both for fixed and for . We prove consistency for and limit theorem for under appropriate assumptions. We obtain both Gaussian and non-Gaussian (stable) limit depending on the growth rate of the power sequence . Applied to real data we find that for larger the estimator is less sensitive to the change in than the Hill estimator.
中文翻译:
极值样本范数的极限律
表示 , 在哪里 , 是一个整数序列,使得 其他 , 和 是 iid 随机变量的阶次统计 索引的上尾有规律地变化 . 估算器是希尔估计量的扩展。我们研究了渐近性质 其他 两者都是固定的 并为 . 我们证明一致性 和极限定理 在适当的假设下。我们根据幂序列的增长率获得高斯和非高斯(稳定)极限. 应用于实际数据,我们发现对于更大的 估计量对变化不太敏感 比 Hill 估计量。