Error Function error function [Gamarnik,Katz,Misra 12]: two distributions ul,u2 :>0,1 E(1,u2）=max -log 1(g) x,y∈2 2(y) marginal distributions u()=Pr[c(v)=xo]and E(u,u)≤exp(-2(t) Pr[c(v)=xo]-Pr[c(v)=x]=exp(-2(t))Error Function error function [Gamarnik, Katz, Misra 12]: two distributions µ1, µ2 : ⌦ ! [0, 1] marginal distributions µ￾ v (x) = Pr[c(v) = x | ￾] and µ⌧ v |Pr[c(v) = x | ￾] ￾ Pr[c(v) = x | ⌧ ]| = exp (￾⌦(t)) E(µ￾ v , µ⌧ v )  exp(￾⌦(t)) E(µ1, µ2) = max x,y2⌦ ✓ log µ1(x) µ2(x) ￾ log µ1(y) µ2(y) ◆