A finite collection of random variables, X<sub>1</sub>,…, X<sub>n</sub> (n≥2), is said to be negatively associated (NA) if any two coordinatewise nondecreasing (or nonincreasing) functi...A finite collection of random variables, X<sub>1</sub>,…, X<sub>n</sub> (n≥2), is said to be negatively associated (NA) if any two coordinatewise nondecreasing (or nonincreasing) functions f<sub>1</sub> and f<sub>2</sub> on R<sup>n</sup>, such that (?)<sub>j</sub>=f<sub>j</sub>(X<sub>1</sub>,…,X<sub>n</sub>) have a finite variance for j=1, 2 and Cov((?)<sub>1</sub>,(?)<sub>2</sub>)≤0; an infinite collection is said to be NA if every finite subcollection is NA. It is relative to a lot of practical problems, such as reliability theory, percolation models and multivariate statistical analysis. Matula (1992) studied the strong laws of large numbers for an NA sequence. Recent-展开更多
文摘A finite collection of random variables, X<sub>1</sub>,…, X<sub>n</sub> (n≥2), is said to be negatively associated (NA) if any two coordinatewise nondecreasing (or nonincreasing) functions f<sub>1</sub> and f<sub>2</sub> on R<sup>n</sup>, such that (?)<sub>j</sub>=f<sub>j</sub>(X<sub>1</sub>,…,X<sub>n</sub>) have a finite variance for j=1, 2 and Cov((?)<sub>1</sub>,(?)<sub>2</sub>)≤0; an infinite collection is said to be NA if every finite subcollection is NA. It is relative to a lot of practical problems, such as reliability theory, percolation models and multivariate statistical analysis. Matula (1992) studied the strong laws of large numbers for an NA sequence. Recent-