Math 171B: Mathematical Programming Homework Assignment #5, Assignments of Mathematics

Mathematical programming homework assignment for math 171b course taught by jennifer erway during spring quarter 2007. The assignment includes various exercises on quadratic functions, stationary points, and minimizers. Some exercises require the use of matlab.

Typology: Assignments

Pre 2010

Uploaded on 03/28/2010

koofers-user-hzl
koofers-user-hzl 🇺🇸

10 documents

1 / 1

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Math 171B: Mathematical Programming
Instructor: Jennifer Erway
Spring Quarter 2007
Homework Assignment #5
Due: Friday, May 4
The starred exercises are those that require the use of MATLAB Remember, you MUST do the MATLAB
problems to get credit for the homework.
Exercise 5.1 Let q(x), x Rn, be the quadratic function q(x) = cTx+1
2xTHx, where His symmetric.
(a). Write down an expression for q(x) in terms of c, H and x.
(b). Given an arbitrary point x0and a direction p, write down the Taylor-series expansion of q(x0+p).
(c). For this part, consider q(x) such that His positive definite. If pis a direction such that q(x0)Tp < 0,
show that there exists a positive minimizer αof q(x0+αp). Derive a closed-form expression for α.
Exercise 5.2 Let f(x) denote a convex continuously differentiable function. Show that if a stationary point
xexists then f(x) is the unique global minimum of f. (Hint: Use the result that f(x) is convex if and
only if f(y)f(x) + f0(x)(yx) for all xand y).
Exercise 5.3 Write down the function value, gradient vector, and Hessian matrix of the function f(x) =
7x1x2+ 3x2
1x2
2at the point (1,1)T. If a stationary point exists, find its position and state the nature
of the stationarity. Does f(x) have a bounded minimizer?
Exercise 5.4* Let q(x) be a function of the form q(x) = cTx+1
2xTHx, where cis a vector and His
a constant symmetric matrix. In each of the problems given below, (i) determine if a stationary point x
of q(x) exists; (ii) if xexists, compute its value and find if it is a minimizer of q; (iii) if xis a minimizer
of q(x), find the minimum value of q; (iv) if xis a stationary point that is not a minimizer of q, find a
direction psuch that q(x+αp)q(x), for all α0.
(a)
c=
1
3
7
1
, H =
4443
4733
4353
3333
(b)
c=
1
1
1
1
, H =
6420
4733
2353
0333
(c)
c=
1
1
2
, H =
213
132
324
(d)
c=
0
1
5
5
, H =
11 7 3 2
7 6 5 2
3 5 7 6
2267
1

Partial preview of the text

Download Math 171B: Mathematical Programming Homework Assignment #5 and more Assignments Mathematics in PDF only on Docsity!

Math 171B: Mathematical Programming

Instructor: Jennifer Erway Spring Quarter 2007

Homework Assignment # Due: Friday, May 4

The starred exercises are those that require the use of MATLAB Remember, you MUST do the MATLAB problems to get credit for the homework.

Exercise 5.1 Let q(x), x ∈ Rn, be the quadratic function q(x) = cTx + 12 xTHx, where H is symmetric.

(a). Write down an expression for ∇q(x) in terms of c, H and x.

(b). Given an arbitrary point x 0 and a direction p, write down the Taylor-series expansion of q(x 0 + p).

(c). For this part, consider q(x) such that H is positive definite. If p is a direction such that ∇q(x 0 )T^ p < 0, show that there exists a positive minimizer α∗^ of q(x 0 + αp). Derive a closed-form expression for α∗.

Exercise 5.2 Let f (x) denote a convex continuously differentiable function. Show that if a stationary point x∗^ exists then f (x∗) is the unique global minimum of f. (Hint: Use the result that f (x) is convex if and only if f (y) ≥ f (x) + f ′(x)(y − x) for all x and y).

Exercise 5.3 Write down the function value, gradient vector, and Hessian matrix of the function f (x) = 7 x 1 − x 2 + 3x^21 − x^22 at the point (− 1 , −1)T. If a stationary point exists, find its position and state the nature of the stationarity. Does f (x) have a bounded minimizer?

Exercise 5.4* Let q(x) be a function of the form q(x) = cTx + 12 xTHx, where c is a vector and H is a constant symmetric matrix. In each of the problems given below, (i) determine if a stationary point x∗ of q(x) exists; (ii) if x∗^ exists, compute its value and find if it is a minimizer of q; (iii) if x∗^ is a minimizer of q(x), find the minimum value of q; (iv) if x∗^ is a stationary point that is not a minimizer of q, find a direction p such that q(x∗^ + αp) ≤ q(x∗), for all α ≥ 0.

(a)

c =

 ,^ H^ =

(b)

c =

 ,^ H^ =

(c)

c =

 , H =

(d)

c =

 ,^ H^ =