Scribd
Upload
18 views41 pages

16 LP 5 Formulations

The document discusses different formulations of linear programming problems and how algorithms that can solve one formulation can be used to solve the others. It explains that full optimization, optimization from a starting point, solution finding, satisfiability, and equivalence are all different problem types that fall under linear programming. The key points are that full optimization can solve all other versions, and the formulations are equivalent because the algorithms for one can be used to solve the others.

Uploaded by

quangnhat1220
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
18 views41 pages

16 LP 5 Formulations

The document discusses different formulations of linear programming problems and how algorithms that can solve one formulation can be used to solve the others. It explains that full optimization, optimization from a starting point, solution finding, satisfiability, and equivalence are all different problem types that fall under linear programming. The key points are that full optimization can solve all other versions, and the formulations are equivalent because the algorithms for one can be used to solve the others.

Uploaded by

quangnhat1220
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/ 41

Linear Programming:

Linear Programming
Formulations
Daniel Kane
Department of Computer Science and Engineering
University of California, San Diego

Advanced Algorithms and Complexity


Data Structures and Algorithms
Learning Objectives
Distinguish between the different types
of linear programming problems.
Use an algorithm that solves one
formulation to solve another
formulation.
Formulations

Several different problem types that all go


under the heading of “linear programming”.
Full Optimization
Minimize or maximize a linear function
subject to a system of linear inequality
constraints (or say that the constraints have
no solution).
Optimization from Starting Point
Given a system of linear inequalities and a
vertex of the polytope they define, optimize
a linear function with respect to these
constraints.
Solution Finding

Given a system of linear inequalities, find


some solution.
Satisfiability

Given a system of linear inequalities


determine whether or not there is a solution.
Equivalence

Actually, if you can solve any of these


problems, you can solve any other!
Full Optimization

Clearly capable of solving all the other


versions.
Start Opt: Ignore starting point.
Solution Finding: Optimal is a solution.
Satisfiability: See if finds a solution.
Optimization from Starting Point

How do you find starting point?


Optimization from Starting Point

How do you find starting point?


Add equations one at a time.
Optimize left hand side of next equation.
Example
Example
Example
Example
Example
Example
Example
Example
Example
Example
Technical Point

Things are a bit messier if some intermediate


systems don’t have optima.
Technical Point

Things are a bit messier if some intermediate


systems don’t have optima.
Fix: start with n constraints (gives a single
vertex). Then while trying to add constraint
v · x ≥ t, don’t just maximize v · x. Also add
v · x ≤ t as a constraint (so that maximum
will exist).
Solution Finding

Q: How do we go from being able to find a


solution to finding the best one?
Solution Finding

Q: How do we go from being able to find a


solution to finding the best one?
A: Duality.
Solution Finding

Q: How do we go from being able to find a


solution to finding the best one?
A: Duality. Find a solution and a matching
dual solution.
Setup
Want to minimize x · v subject to Ax ≥ b.
Setup
Want to minimize x · v subject to Ax ≥ b.
Instead find solution to:

Ax ≥ b
y ≥0
yTA = v
x · v = y · b.

Will give optimal solution to original problem.


Satisfiability

How does just knowing when you have a


solution help?
Satisfiability

How does just knowing when you have a


solution help?
Can always find solution at a vertex. Means
n equations are tight.
Satisfiability

How does just knowing when you have a


solution help?
Can always find solution at a vertex. Means
n equations are tight.
Figure out which equations to use.
Example
Example
Example
Example
Example
Example
Example
Example
Problem

In order to find a solution to a linear program


with m equations in n variables, how many
times would one have to call a satisfiability
algorithm?
Solution

In order to find a solution to a linear program


with m equations in n variables, how many
times would one have to call a satisfiability
algorithm?
m times. You need to test each equation
once, keeping the ones that work.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy