IE 5531: Engineering Optimization I

Similar documents
Introduction to Mathematical Programming IE406. Lecture 10. Dr. Ted Ralphs

Linear Programming Duality P&S Chapter 3 Last Revised Nov 1, 2004

Linear Programming Duality

Note 3: LP Duality. If the primal problem (P) in the canonical form is min Z = n (1) then the dual problem (D) in the canonical form is max W = m (2)

IE 5531: Engineering Optimization I

4.6 Linear Programming duality

Lecture 10: Linear programming. duality. and. The dual of the LP in standard form. maximize w = b T y (D) subject to A T y c, minimize z = c T x (P)

Algorithms and Theory of Computation. Lecture 13: Linear Programming (2)

Discrete Optimization

F 1 F 2 Daily Requirement Cost N N N

Optimisation and Operations Research

Lecture 5. x 1,x 2,x 3 0 (1)

Summary of the simplex method

II. Analysis of Linear Programming Solutions

Linear Programming Inverse Projection Theory Chapter 3

Sensitivity Analysis and Duality

Lectures 6, 7 and part of 8

Sensitivity Analysis and Duality in LP

(P ) Minimize 4x 1 + 6x 2 + 5x 3 s.t. 2x 1 3x 3 3 3x 2 2x 3 6

Chapter 1 Linear Programming. Paragraph 5 Duality

Example Problem. Linear Program (standard form) CSCI5654 (Linear Programming, Fall 2013) Lecture-7. Duality

Farkas Lemma, Dual Simplex and Sensitivity Analysis

Lecture 10: Linear programming duality and sensitivity 0-0

MVE165/MMG631 Linear and integer optimization with applications Lecture 5 Linear programming duality and sensitivity analysis

CSCI5654 (Linear Programming, Fall 2013) Lecture-8. Lecture 8 Slide# 1


Linear and Combinatorial Optimization

Lecture 11: Post-Optimal Analysis. September 23, 2009

Part 1. The Review of Linear Programming

Chap6 Duality Theory and Sensitivity Analysis

IE 5531: Engineering Optimization I

Duality Theory, Optimality Conditions

ECE 307- Techniques for Engineering Decisions

Review Solutions, Exam 2, Operations Research

Introduction to Mathematical Programming

MAT016: Optimization

BBM402-Lecture 20: LP Duality

The use of shadow price is an example of sensitivity analysis. Duality theory can be applied to do other kind of sensitivity analysis:

IE 5531: Engineering Optimization I

CS261: A Second Course in Algorithms Lecture #9: Linear Programming Duality (Part 2)

LP Duality: outline. Duality theory for Linear Programming. alternatives. optimization I Idea: polyhedra

Linear Programming: Chapter 5 Duality

4. Duality and Sensitivity

Midterm Review. Yinyu Ye Department of Management Science and Engineering Stanford University Stanford, CA 94305, U.S.A.

Spring 2017 CO 250 Course Notes TABLE OF CONTENTS. richardwu.ca. CO 250 Course Notes. Introduction to Optimization

The Primal-Dual Algorithm P&S Chapter 5 Last Revised October 30, 2006

The Simplex Algorithm

Review Questions, Final Exam

LINEAR PROGRAMMING II

4. Duality Duality 4.1 Duality of LPs and the duality theorem. min c T x x R n, c R n. s.t. ai Tx = b i i M a i R n

CO 250 Final Exam Guide

CS261: A Second Course in Algorithms Lecture #8: Linear Programming Duality (Part 1)

SEN301 OPERATIONS RESEARCH I LECTURE NOTES

1 Review Session. 1.1 Lecture 2

MS-E2140. Lecture 1. (course book chapters )

SAMPLE QUESTIONS. b = (30, 20, 40, 10, 50) T, c = (650, 1000, 1350, 1600, 1900) T.

1. Algebraic and geometric treatments Consider an LP problem in the standard form. x 0. Solutions to the system of linear equations

MS-E2140. Lecture 1. (course book chapters )

Introduction to linear programming using LEGO.

The Dual Simplex Algorithm

Duality of LPs and Applications

Introduction to Operations Research Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras

x 1 + 4x 2 = 5, 7x 1 + 5x 2 + 2x 3 4,

IE 5531: Engineering Optimization I

Review Questions, Final Exam

Minimum cost transportation problem

How to Take the Dual of a Linear Program

4. The Dual Simplex Method

IE 5531: Engineering Optimization I

A Review of Linear Programming

Linear programming. Saad Mneimneh. maximize x 1 + x 2 subject to 4x 1 x 2 8 2x 1 + x x 1 2x 2 2

Answer the following questions: Q1: Choose the correct answer ( 20 Points ):

Fundamentals of Operations Research. Prof. G. Srinivasan. Indian Institute of Technology Madras. Lecture No. # 15

"SYMMETRIC" PRIMAL-DUAL PAIR

Relation of Pure Minimum Cost Flow Model to Linear Programming

Part IB Optimisation

Lecture 15 (Oct 6): LP Duality

END3033 Operations Research I Sensitivity Analysis & Duality. to accompany Operations Research: Applications and Algorithms Fatih Cavdur

56:171 Operations Research Fall 1998

Lec12p1, ORF363/COS323. The idea behind duality. This lecture

6.854J / J Advanced Algorithms Fall 2008

The dual simplex method with bounds

System Planning Lecture 7, F7: Optimization

3. Duality: What is duality? Why does it matter? Sensitivity through duality.

COT 6936: Topics in Algorithms! Giri Narasimhan. ECS 254A / EC 2443; Phone: x3748

IE 5531: Engineering Optimization I

UNIT-4 Chapter6 Linear Programming

Another max flow application: baseball

ORIE 6300 Mathematical Programming I August 25, Lecture 2

Linear Programming. Jie Wang. University of Massachusetts Lowell Department of Computer Science. J. Wang (UMass Lowell) Linear Programming 1 / 47

MATH 445/545 Test 1 Spring 2016

56:270 Final Exam - May

The Strong Duality Theorem 1

Topic: Primal-Dual Algorithms Date: We finished our discussion of randomized rounding and began talking about LP Duality.

Planning and Optimization

Linear and Integer Programming - ideas

min 4x 1 5x 2 + 3x 3 s.t. x 1 + 2x 2 + x 3 = 10 x 1 x 2 6 x 1 + 3x 2 + x 3 14

Week 3 Linear programming duality

Metode Kuantitatif Bisnis. Week 4 Linear Programming Simplex Method - Minimize

2.098/6.255/ Optimization Methods Practice True/False Questions

Transcription:

IE 5531: Engineering Optimization I Lecture 7: Duality and applications Prof. John Gunnar Carlsson September 29, 2010 Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 1 / 30

Administrivia PS 2 posted this evening Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 2 / 30

Recap Any linear program minimize c T x s.t. Ax = b x 0 has a dual of the form maximize b T y s.t. A T y c If both problems are feasible, we have c T x b T y for all feasible x and y At optimality, we have c T x = b T y The dual of the dual is the primal Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 3 / 30

Duality relationships Primal minimize c T x maximize b T y Dual A A T b i 0 constraints b i 0 variables = b i free 0 c j variables 0 c j constraints free = c j Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 4 / 30

Today Weak & strong duality Duality examples Complementary slackness Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 5 / 30

Weak duality The weak duality theorem says that if x is a feasible point for the primal (minimization) problem and y is a feasible point for the dual (maximization) problem, then we have b T y c T x If optimal solutions exist for both problems, the strong duality theorem says that b T y = c T x Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 6 / 30

Proof of weak duality Let x and y be feasible primal-dual pairs for a standard-form LP, so that Ax = b; x 0; A T y c We nd that b T y = x T A T y c T x since x 0 Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 7 / 30

Proof of strong duality Suppose that x is an optimal solution to a standard-form LP that we found via the simplex method, let B denote the basic indices at x, and let N denote the non-basic indices Let x B = A 1 B b and x N = 0 Recall the denition of the reduced cost vector r in the simplex method (the top row): r = c A ( ) T A 1 T B c B When the simplex method terminates, the reduced cost vector must be nonnegative, thus ) T c B 0 c A ( T A 1 B }{{} y Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 8 / 30

Proof of strong duality Dene y by setting y = ( ) A 1 T B c B as in the preceding; clearly we have and thus y is dual feasible Finally, note that b T y = b T ( A 1 B which completes the proof c A T y ) T c B = ( ) A 1 T b c B = c T B x = B ct x y must be optimal for the dual problem, by weak duality B Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 9 / 30

Unboundedness Note that if one of the problems is unbounded, then the other must be infeasible Suppose that c T x is unbounded below in a standard-form LP; weak duality says that for all x and y b T y c T x Suppose that y is feasible in the dual; we can choose x to make c T x arbitrarily small and thus b T y, a contradiction Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 10 / 30

Infeasibility On the other hand, it is not true that if one problem is infeasible, the other must be unbounded The following primal-dual pairs are both infeasible: with dual minimize x 1 + 2x 2 x 1 + x 2 = 1 2x 1 + 2x 2 = 3 maximize y 1 + 3y 2 y 1 + 2y 2 = 1 y 1 + 2y 2 = 2 s.t. s.t. Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 11 / 30

Feasibility and boundedness An obvious corollary: if the primal problem is infeasible and the dual problem has a feasible point, then it must be unbounded A less obvious corollary: unless both problems are infeasible, at least one of them must have an unbounded feasible set The following relationships between primal and dual LPs are possible: Primal Finite optimum Unbounded Infeasible Finite optimum Possible X X Dual Unbounded X X Possible Infeasible X Possible Possible Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 12 / 30

Duality example Consider a transportation problem: We have m warehouses, each containing an amount a i of a product We have n stores that need to receive an amount b j of that product It costs c ij to transport one unit of the product from warehouse i to store j How can we minimize our costs while satisfying all stores' demands? Assume that m i=1 a i = n j=1 b j (total supply equals total demand) Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 13 / 30

Duality example The LP formulation is minimize m n c ij x ij s.t. i=1 j=1 m x ij = a i i j=1 n x ij = b j j i=1 x ij 0 i, j Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 14 / 30

Duality example The dual of the transportation problem is maximize m n a i u i + b j v j s.t. i=1 j=1 u i + v j c ij i, j Remember: ai is the amount of the product available at warehouse i bj is the amount of the product required at store j cij is the transportation cost from i to j Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 15 / 30

Duality example Since there is no sign constraint, we can set u i u i (for interpretation purposes): n m maximize b j v j a i u i s.t. j=1 i=1 v j u i c ij i, j Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 16 / 30

Duality example Since there is no sign constraint, we can set u i u i (for interpretation purposes): n m maximize b j v j a i u i s.t. j=1 i=1 v j u i c ij i, j Interpretation: think of the u i 's and v j 's as prices Specically, think of a new logistics company that wants to ship these products The company will buy the goods at each warehouse at a price u i and sell them back to the warehouses at a price v j In order to be competitive with the original company's transportation modes, the new company must have v j u i c ij i, j (otherwise the original transportation company won't use them) Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 16 / 30

Maximum ow Recall from earlier the maximum ow problem: Given a directed, weighted graph G = (V, E, W ) and a pair of nodes s and t Think of the edge weights w ij as the capacity of that edge What's the largest amount of ow we can send from s to t, subject to the capacity constraints? Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 17 / 30

Maximum ow In lecture 1 we derived an LP formulation: let x ij denote the amount of ow across edge (i, j); the problem is maximize (s,i) E (i,j) E x si s.t. x ij w ij (i, j) E x ij = x ji i s, t (j,i) E x ij 0 (i, j) E Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 18 / 30

Maximum ow The dual of the maximum ow problem is minimize w ij z ij s.t. (i,j) E z ij = y i y j (i, j) E y s y t = 1 z ij 0 (i, j) E (see B&T 7.5 for derivation) Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 19 / 30

Maximum ow The dual of the maximum ow problem is minimize w ij z ij s.t. (i,j) E (see B&T 7.5 for derivation) z ij = y i y j (i, j) E y s y t = 1 z ij 0 (i, j) E We're assigning values y i to the nodes, and incurring costs y i y j for all (i, j) E We can assume WLOG that y s = 1 and y t = 0 It is not hard to show that, in fact, y i {0, 1} for all i Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 19 / 30

Min cut The preceding problem is equivalent to nding a subset S of vertices, such that s S, that minimizes the cut length of S, i.e. i S:j / S w ij Thus, strong duality says that the maximum ow we can push through the network is equal to the smallest cut size of any subset S of vertices Corollary: if the value of a ow is equal to the value of some cut, then both are optimal Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 20 / 30

Max-ow min-cut theorem Input graph Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 21 / 30

Max-ow min-cut theorem Maximum ow Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 22 / 30

Max-ow min-cut theorem Minimum cut Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 23 / 30

Complementary slackness Consider a standard-form linear program and let A i denote column i of A We know that any constraint in the primal corresponds to a variable in the dual, and vice versa An important relationship between optimal primal and dual solutions is known as complementary slackness, or complementarity: Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 24 / 30

Complementary slackness Theorem Let x and y be feasible solutions to the primal and dual problem, respectively. The vectors x and y are optimal solutions for the two respective problems if and only if x i > 0 = A T i y = c i A T i y < c i = x i = 0 in other words, x i (A T i y c i ) = 0 i Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 25 / 30

Proof of complementary slackness Suppose that A T i x i > 0 = A T i y = c i y < c i = x i = 0 as in the theorem; then certainly ( A T y c) T x = = = 0 (y T A c T ) n i=1 x (y T A i c i ) x i Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 26 / 30

Proof of complementary slackness Since ( y T A c T ) x = 0, we know that 0 = (y T A c T ) = y T Ax c T x = y T b c T x c T x = b T y x which completes one direction of the proof by weak duality. Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 27 / 30

Proof of complementary slackness To prove the other direction, suppose that x and y are optimal, so that c T x = b T y (strong duality) We can write c T x b T y = 0 c T x (Ax) T y = 0 c T x y T Ax = 0 ( ) c T y T A x = 0 Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 28 / 30

Proof of complementary slackness We can rewrite the equation ( ) c T y T A x = 0 as a summation: n i=1 Since x and y are feasible, we have so ( c i y T A i ) xi for all i (c i y T A i ) x i = 0 c i y T A i 0 i x i 0 i Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 29 / 30

Non-standard form LPs Complementarity holds for LPs that are not in standard form as well Let a T j denote the jth row of A and let A i denote the ith column of A A primal-dual pair x, y is optimal if and only if ) y j (a T j x b j = 0 j {1,..., m} ) (c i y T A i x i = 0 i {1,..., n} where A R m n In other words: take each variable, multiply it by its associated constraint; the product must be 0 Prof. John Gunnar Carlsson IE 5531: Engineering Optimization I September 29, 2010 30 / 30