A Comparison Inequality for Positively Associated Random Variables
Wednesday, April 29, 2015 - 3:10pm - 4:00pm
We present a general comparison inequality between positively associated and independently distributed random variables, generalizing the Gaussian comparison inequality. The proof uses several combinatorial identities and the property of positive association. We also provide applications where the maxima of correlated r.v.s behaves in the same way as maxima of i.i.d. r.v.s.