Scribd
Upload
89 views5 pages

IE426 - Optimization Models and Applications: 1 Reformulate As A Linear Programming Problem (8pts.)

This document is a quiz for an optimization modeling course. It contains 4 questions assessing skills in reformulating problems as linear programs, goal programming, integer programming, and branch and bound algorithms. Students are asked to reformulate convex but non-linear objective functions as linear programs, write goal programs to minimize constraint violations, use integer variables to model logical relationships, and describe the node structure and pruning process of branch and bound trees. The quiz is worth 10% of the student's final grade and must be completed within 75 minutes.

Uploaded by

lynndong0214
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
89 views5 pages

IE426 - Optimization Models and Applications: 1 Reformulate As A Linear Programming Problem (8pts.)

This document is a quiz for an optimization modeling course. It contains 4 questions assessing skills in reformulating problems as linear programs, goal programming, integer programming, and branch and bound algorithms. Students are asked to reformulate convex but non-linear objective functions as linear programs, write goal programs to minimize constraint violations, use integer variables to model logical relationships, and describe the node structure and pruning process of branch and bound trees. The quiz is worth 10% of the student's final grade and must be completed within 75 minutes.

Uploaded by

lynndong0214
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

IE426 Optimization models and applications

Fall 2012 Quiz #2, November 15, 2012


First name
Last name
Lehigh email
You have 75 minutes. This quiz accounts for 10% of the final grade. There
are 40 points available. Please write clear and concise statements. Use a
readable handwriting: it will be hard to grade your answers if they are not
readable let alone give them full points. Use the back of each sheet if
you need more space, for example to draw graphs. Do not use calculators.
For each model, clearly specify the meaning of each variable and of each
constraint.

Reformulate as a Linear Programming Problem (8pts.)

Consider the following problem, where the vectors c1 , . . . , ck , b and the matrix A are
given:
Pk
i T
min
i=1 |(c ) x|
st.
Ax b
The objective function is convex but not linear, since it is the sum of absolute values.
Reformulate this problem as a linear programming problem using techniques studied
in class introducing appropriate extra variables.

Mixed Integer/Goal Programming (14 pts.)

The following LP is infeasible


min

3x1
x1
2x1
x1
x1

+5x2
+2x2
+x2
x2
x2

2
1
4
[0, 5]
[0, 5]

1. Write a nonpreemptive goal programming formulation to minimize the constraint


violation. Do not solve! (6pts)

2. In nonpreemptive goal programming we minimize the total sum of the constraint


violation. You can think of it as minimizing the cost of violating constraints if
each time you violate a constraint by one unit you pay one unit of cost. Now
imagine that each time you violate one of the constraints you pay a fixed cost of
20 on top of the per-unit cost of 1. Notice that for the fixed cost we do not care
by how much these constraints are violated, as long as they are. Formulate the
problem of minimizing the total cost of violating the constraints. (6 pts).

(contd)

3. Write the good values for the big M constants (2pts).

Integer Programming and Logic (12pts)

Formulate the following sets of constraints via mixed integer programming


1. max{(a1 x1 + a2 x2 + . . . , an xn ), (b1 x1 + b2 x2 + . . . , bn xn )} 0.
2. If a1 x1 + a2 x2 + . . . , an xn 0 then y = 1, for binary y.
Pn
3.
i=1 xi = 1, 0 x 1, at most k elements of x are allowed to be positive for
some given k < n.

Branch and Bound (6 pts)

Briefly but accurately give answers to the following questions.


1. The process of a Branch and Bound algorithm is represented by a tree. What are
the nodes of that tree? What does each node represent and how nodes that are
connected by an edge are related to each other?
2. What is the largest number of nodes and a smallest number of nodes a B&B tree
can have for a problem with n binary variables.
3. What should happen for the tree to have the smallest number of nodes?

You might also like