Probability with Engineering Applications: Problem Set 1 Solutions, Assignments of Statistics

Material Type: Assignment; Class: Probability with Engrg Applic; Subject: Electrical and Computer Engr; University: University of Illinois - Urbana-Champaign; Term: Fall 2002;

Typology: Assignments

Pre 2010

Uploaded on 03/16/2009

koofers-user-6sf
koofers-user-6sf 🇺🇸

10 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
ECE/CS 313: Probability with Engineering Applications Fall 2002
Problem Set 1 Solutions
1. Geometric series
(a) Since x6= 1, the equation to be proved is equivalent to the equation we get by multiplying both sides by
(1 x). So it is sufficient to prove that (1 x)(1 + x+x2+...+xn1) = 1 xn. But
(1 x)(1 + x+x2+...+xn1) = (1 + x+x2+...+xn1)x(1 + x+x2+...+xn1)
= (1 + x+x2+...+xn1)(x+x2+x3...+xn)
= 1 xn,
which establishes the desired equation. (Another way to prove the equation is by induction on n.)
(b) Since limn→∞ xn= 0 under the assumption |x|<1, taking the limit as n in both sides of the
equation in part (a) yields that 1 + x+x2+...=1
1x.
2. Inverse function
Sketching f, we see that fis monotonically decreasing, with limit 2 at −∞ and limit 0 at . In particular,
an inverse function exists since fis one-to-one. The range of fis (0,2), so the inverse gneeds only to
be defined on that interval. We need f(g(u)) = ufor 0 <u<2. Solving this equation for g(u) yields
2
eg(u)+1 =u, or 2
u1 = eg(u), or g(u) = ln( 2
u1) for 0 <u<2.
3. Extrema of functions
(a) The function f(x) = x4(1.001)x2can be written in the form xnexp(ax2), where a= ln(1.001), n = 4.
The function xnexp(ax2) has local extrema at the points x=±∞,0,±pn/2a. In this case n= 4 and
therefore the maxima occur at x=±pn/2a=±44.7325 (as can be verified by differentiating once again and
checking the sign of the result at the two points); the points x= 0,±∞ represent minima. The maximum
function value (at both maxima) is 5.4188 ×105.
(b) If f(x) = x4(1.001)x2, the function converges to infinity as |x| . Thus, f(x) is not a maximum value
of ffor any finite value of x. Incidentally, carrying out the differentiation gives us x=±p2/ln(1.001) as
the other solutions, which is not possible since the quantity within the sign is negative.
4. Some definite integrals
(a) Drawing a picture, you see that the integral from 1 to 1 is just the sum of the areas of two triangles,
each having height 1 and base length 1. Therefore, it is easy to see that the total area = 2 ×1/2 = 1.
Try and draw a little sketch before you tackle most problems. Those who used formal integration principles
should note that |x|=xif x0 and |x|=xif x < 0. Therefore
Z1
1|x|dx =Z0
1x dx +Z1
0
x dx =x2
2
0
1
+x2
2
1
0
= 1.
(b) Substitute y= (1 x2). Then
Z1
0
x(1 x2)11 dx =1
2Z0
1
y11 dy =1
2Z1
0
y11 dy =1
24.
(c) Integrate by parts twice:
Z1
0
x2exdx =x2ex|1
0+Z1
0
2xexdx
=e1+Z1
0
2xexdx
=e12xex|1
0+Z1
0
2exdx
=3e1+Z1
0
2exdx
=3e1+ 2(1 e1) = 2 5e10.1602
pf2

Partial preview of the text

Download Probability with Engineering Applications: Problem Set 1 Solutions and more Assignments Statistics in PDF only on Docsity!

ECE/CS 313: Probability with Engineering Applications Fall 2002

Problem Set 1 Solutions

  1. Geometric series (a) Since x 6 = 1, the equation to be proved is equivalent to the equation we get by multiplying both sides by (1 − x). So it is sufficient to prove that (1 − x)(1 + x + x^2 +... + xn−^1 ) = 1 − xn. But

(1 − x)(1 + x + x^2 +... + xn−^1 ) = (1 + x + x^2 +... + xn−^1 ) − x(1 + x + x^2 +... + xn−^1 ) = (1 + x + x^2 +... + xn−^1 ) − (x + x^2 + x^3... + xn) = 1 − xn,

which establishes the desired equation. (Another way to prove the equation is by induction on n.) (b) Since limn→∞ xn^ = 0 under the assumption |x| < 1, taking the limit as n → ∞ in both sides of the equation in part (a) yields that 1 + x + x^2 +... = (^1) −^1 x.

  1. Inverse function Sketching f , we see that f is monotonically decreasing, with limit 2 at −∞ and limit 0 at ∞. In particular, an inverse function exists since f is one-to-one. The range of f is (0, 2), so the inverse g needs only to be defined on that interval. We need f (g(u)) = u for 0 < u < 2. Solving this equation for g(u) yields 2 eg(u)+1 =^ u, or^

2 u −^ 1 =^ e

g(u), or g(u) = ln( 2 u −^ 1) for 0^ < u <^ 2.

  1. Extrema of functions (a) The function f (x) = x^4 (1.001)−x

2 can be written in the form xn^ exp(−ax^2 ), where a = ln(1.001), n = 4.

The function xn^ exp(−ax^2 ) has local extrema at the points x = ±∞, 0 , ±

n/ 2 a. In this case n = 4 and

therefore the maxima occur at x = ±

n/ 2 a = ± 44 .7325 (as can be verified by differentiating once again and checking the sign of the result at the two points); the points x = 0, ±∞ represent minima. The maximum function value (at both maxima) is 5. 4188 × 105.

(b) If f (x) = x^4 (1.001)x

2 , the function converges to infinity as |x| → ∞. Thus, f (x) is not a maximum value

of f for any finite value of x. Incidentally, carrying out the differentiation gives us x = ±

− 2 / ln(1.001) as the other solutions, which is not possible since the quantity within the “

” sign is negative.

  1. Some definite integrals (a) Drawing a picture, you see that the integral from −1 to 1 is just the sum of the areas of two triangles, each having height 1 and base length 1. Therefore, it is easy to see that the total area = 2 × 1 /2 = 1. Try and draw a little sketch before you tackle most problems. Those who used formal integration principles should note that |x| = x if x ≥ 0 and |x| = −x if x < 0. Therefore

∫ (^1)

− 1

|x| dx =

− 1

−x dx +

0

x dx =

−x^2 2

0

− 1

x^2 2

1

0

(b) Substitute y = (1 − x^2 ). Then ∫ (^1)

0

x(1 − x^2 )^11 dx = −

1

y^11 dy =

0

y^11 dy =

(c) Integrate by parts twice: ∫ (^1)

0

x^2 e−xdx = −x^2 e−x|^10 +

0

2 xe−xdx

= −e−^1 +

0

2 xe−xdx

= −e−^1 − 2 xe−x|^10 +

0

2 e−xdx

= − 3 e−^1 +

0

2 e−xdx

= − 3 e−^1 + 2(1 − e−^1 ) = 2 − 5 e−^1 ≈ 0. 1602

(d) Notice that the integrand is an odd function and therefore it has to integrate to 0 over any interval that is symmetric about the origin. In particular, the integral over the whole real line is zero.

  1. Derivatives and integrals (a) True as per chain rule. (b) Also true as per chain rule. (c) False. Chain rule gives (^) dxd exp

f (x^2 )

as 2x · g(x^2 ) exp

f (x^2 )

(d) False.

g(−x) dx = −f (−x) + C (minus sign missing.) (e) False. The antiderivative of g(x^2 /2) need not be related to f (x^2 /2) at all, as can be seen by differentiating both sides of (v). For example, take f (x) = ln(1 + x) to see this. (f) True only for positive functions f (x).

  1. Double integrals (a) f (x, y) = max(x, y), over the region D = {(x, y) : 0 ≤ x ≤ 1 , 0 ≤ y ≤ 1 }. Drawing a three-dimensional picture is not the easiest thing, but it is worth the effort to get an idea of what the function looks like. Applying basic integration techniques and using the fact that max(x, y) = y if x ≤ y and max(x, y) = x if x > y, we get

∫ (^1)

0

0

max(x, y) dx dy =

0

(∫ (^) y 0 y dx

dy +

0

1 y x dx

dy

0 y

(^2) dy + ∫^1 0

1 −y^2 2 dy^ =^

1 3 +^

1 3 =^

2

(b) f (x, y) = (x^2 + y^2 )−^4 , over the region D = {(x, y) : x^2 + y^2 > 1 }. Changing to polar coordinates (x = r cos θ, y = r sin θ), the above integral can be reduced to

∫ ∫

D

(x^2 + y^2 )−^4 =

1

∫ (^2) π

0

r−^8 · r drdθ = 2π

1

r−^7 dr = π/ 3 ,

where the two-dimensional differential dx dy transforms to r dr dθ.

  1. An application of Taylor’s approximation (a) f (0) = 0 and f ′(0) = 1, so P (x) = x. (b) |f ′′(x)| = | (^) (1+^1 x) 2 | which is monotonically decreasing on the interval [− 0. 5 , 0 .5]. Thus, the maximum of

|f ′′| over the interval is |f ′′(− 0 .5)| = 4. Therefore, we can take M = 4. (Any larger value of M would also work.) (c) Note that n ln(1 + a n ) = nf ( a n ) ≈ nP ( a n ) = a. That is, the limit is a. To do a more careful job, we can bound the rate of convergence as follows. If n ≥ 2 a then | a n | ≤ 0 .5 so Taylor’s Inequality with M = 4 yields

|n ln(1 +

a n

) − a| = n|f (

a n

) − P (

a n

≤ n

M

a n

)^2 if n ≥ 2 a

2 a^2 n

(d) Using the hint,

a n

)n^ = exp(ln((1 +

a n

)n))

= exp(n ln(1 +

a n

(1) → exp(a)

The limit (1) is justified because the argument of the exp function converges as shown in part (c), and the exp function is continuous.