18.7. Sums of Random Variables 641
Proof.
Ex
h
cT
i
DEx
h
cT^1 CCTn
i
(def ofT)
DEx
h
cT^1 cTn
i
DEx
h
cT^1
i
ExŒcTnç (independent product Cor 17.5.7)
e.c 1/ExŒT^1 çe.c 1/ExŒTnç (by Lemma 18.7.3 below)
De.c 1/.ExŒT^1 çCCExŒTnç/
De.c 1/ExŒT^1 CCTnç (linearity of ExŒç)
De.c 1/ExŒTç:
Lemma 18.7.3.
ExŒcTiçe.c 1/ExŒTiç
Proof. All summations below range over valuesvtaken by the random variableTi,
which are all required to be in the intervalŒ0;1ç.
ExŒcTiçD
X
cvPrŒTiDvç (def of ExŒç)
X
.1C.c 1/v/PrŒTiDvç (convexity —see below)
D
X
PrŒTiDvçC.c 1/vPrŒTiDvç
D
X
PrŒTiDvçC.c 1/
X
vPrŒTiDvç
D 1 C.c 1/ExŒTiç
e.c 1/ExŒTiç (since 1 Czez):
The second step relies on the inequality
cv 1 C.c 1/v;
which holds for allvinŒ0;1çandc 1. This follows from the general principle
that a convex function, namelycv, is less than the linear function, 1 C.c 1/v,
between their points of intersection, namelyvD 0 and 1. This inequality is why
the variablesTiare restricted to the intervalŒ0;1ç.