Limit Theorems for Lower-Upper Extreme Values from Two-Dimensional Distribution Function | ||||
The Egyptian Statistical Journal | ||||
Article 6, Volume 32, Issue 2, December 1988, Page 153-167 | ||||
Document Type: Original Article | ||||
DOI: 10.21608/esju.1988.316570 | ||||
View on SCiNiTO | ||||
Author | ||||
H. M. Barakat | ||||
Abstract | ||||
The limiting distribution of the random vector (V̅_(k,k^':n) - b̅_n) / a̅_n = (X_(1,k:n) - b₁ₙ) / a₁ₙ, (X_(2,n-k^'+1:n) - b₂ₙ) / a₂ₙ k and k' being constants, are investigated, necessary and sufficient conditions for waking convergence of the distribution of the above vector are obtained. The conditions under which the components of the vector (V̅_(k,k^':n) - b̅_n) / a̅_n are asymptotically independent are also obtained. Some cases are examined when the number of observations is a random variable. | ||||
Keywords | ||||
Convergence; Extreme Values; Limit Theorems; Necessary and Sufficient Conditions; Two; Dimensional Distribution Function | ||||
Statistics Article View: 21 |
||||