Let Xi, Xz, Xyr - sequence of independent identically distributed random variables with zero mean and finite variance.
Show that the characteristic function of Sv Yno" XN Xk can be written as: @sn (@) (ai" 4" Ax). The Taylor formula f(x) = f(0) + is valid for very small x. When N in (a) is very large, apply Taylor formula for @x (WNaz) and use the relation between the characteristic function and moments to show that: @sx (@). Use Jin (1+4" e*) to find Jie_ WS, (6).
Parts (a)-(e) prove the Central Limit Theorem.
Parts (d) and (e) use software, but submit.
(d) Let Xy, Xz, Xyr be a sequence of independent identically distributed Uniform random variables on the interval [-1,1]. Write code to verify the Central Limit Theorem by plotting the histogram for S5, S10, Sz0. Use enough sample points (i.e. realization of the sequence X; Xz, X3, ..., Xv) to generate clear plots.
(e) Let X, Xz, Xy be a sequence of independent identically distributed Cauchy random variables on the interval [0,∞). Write code to plot the histogram for S,, S10, Sz0. Does the Central Limit Theorem hold in this case?
Hint: There is built-in code for the Cauchy distribution in MATLAB. Instead, you can use the following code for part (e). Also, the same code can be modified slightly and used for part (d):
Clear N = 5:810,
for I = 1: 1000
(n) = 4Lm
TtLam (pi * (rand(1,N)
51)) {sgrL (N)
end
histogram(C) axis([-1000 1000 500])