Distributionshard
0:00.0

Given X∼Binomial(n,p)X \sim \text{Binomial}(n, p), find the MGF MX(t)=E[etX]M_X(t) = E[e^{tX}].