Duality Principle in Order Statistics
本文在任意情形下建立了顺序统计量的对偶原理，可无需正式证明即推导出许多已知的对偶结果，对研究顺序统计量性质有用。
SUMMARY We establish a duality principle for order statistics in the arbitrary case, using which many known dual results on order statistics can be deduced without a formal proof. The literature on order statistics is rich with such dual pairs derived under different assumptions such as continuity, absolute continuity, exchangeability, etc.