IP Duality. Menal Guzelsoy. Seminar Series, /21-07/28-08/04-08/11. Department of Industrial and Systems Engineering Lehigh University
|
|
- Abel Clarke
- 6 years ago
- Views:
Transcription
1 IP Duality Department of Industrial and Systems Engineering Lehigh University Seminar Series, /21-07/28-08/04-08/11
2 Outline Duality Theorem 1 Duality Theorem Introduction Optimality Conditions 2 Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example 3 Characterization Solution Procedure 4 With an OSF With Branch and Bound Inference Duality 5 Methods to Solve PIPs
3 References: Duality Theorem Diego Klabjan A New Subadditive Approach to Integer Programming: Theory and Algorithms In Proceedings of the Ninth Conference on Integer Programming and Combinatorial Optimization, Cambridge, MA, pages L.A. Wolsey Integer Programming Duality: Price Functions and Sensitivity Analaysis Math Programming, 20 L. Schrage and L.A. Wolsey Sensitivity analysis for branch and bound integer programming. Operations Research, 33(5): , M. Dawande and J. N. Hooker Inference-based sensitivity analysis for mixed integer/linear programming, revised 1997, Operations Research 48 (2000)
4 References: Duality Theorem Jenkins, L. (1990). Parametric Methods in Integer Linear Programming. Annals of Operation Research. 27, Jenkins,L. Parametric-objective integer programming using knapsack facets and Gomory cutting planes. Eur. J. Oper. Res. 31 (1987) Wang, HF & Jyh-Shing Horng, 1996, Structural Approach to Parametric Analysis of an IP, European J. of OR, 92, Marsten RE and Morin TL, Parametric integer programming: the right-hand side case, Annals of. Discr. Math. 1 (1977) Geoffrion A.M. and Nauss R. Parametric and postoptimality analysis in integer.linear programming. Management Science, 23(5): , Rountree S., and B. Gillett, Parametric integer linear programming: A synthesis of branch and bound with cutting planes, European Journal of Operational Research, 10, , 1982
5 Duality Theorem Introduction Optimality Conditions Z = max s.t nx c j x j j=1 nx a j x j b a j, b Z m (P) j=1 x j 0 integer, j = 1,..., n W = min F(b) s.t F(a j ) c j j = 1,..., n (D) F Γ m where Γ m = {F : R m R F is superadditive, nondecreasing and F(0) = 0}.
6 Z=W Duality Theorem Introduction Optimality Conditions Weak Duality: cx F(b) for all primal feasible vectors x and dual feasible functions F Z W (P) is infeasible if W (D) is infeasible if Z Strong Duality: If either (P) or (D) has a finite optimal value, then there exists x and F s.t Z = cx = F (b) = W If (P) is infeasible, either (D) is infeasible or W If (D) is infeasible, either (P) is infeasible or Z
7 Complementary Slackness Duality Theorem Introduction Optimality Conditions Let x and F be optimal solutions to (P) and (D). Then, c j F (a j ) 0, j = 1,..., n If x j > 0, then c j F (a j ) = 0 F is nondecreasing and superadditive F (b) = F ( P n j=1 a jx j ) and F (b P n j=1 a jx j ) = 0
8 Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example What type of dual optimal functions can be generated? What else can we know about the structure of F? Theorem of Alternative: P n j=1 π jx j π 0 is a valid inequality for Q = {x Ax b, x 0, and integer} iff there exists F Γ m s.t F (a j ) π j, j = 1,..., n and F (b) π 0. conv[q] {x nx F(a j )x j F(b), F Γ m, x 0} j=1
9 Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example Consider the problems P 1, D 1, D 2 : Z 1 = max cx nx s.t F 1 (h j )x j F 1 (h) F 1 G (P 1 ) j=1 x conv[x : Ex e, x 0, integer] W 1 = inf F 1 G nx max{ [(c j F 1 (h j ))x j ] + F 1 (h) : Ex e, x 0, integer} (D 1 ) j=1 W 2 = inf {F(b) : F(a j ) c j, F G Γ m r } (D 2 )
10 Propositions Duality Theorem Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example Z Z 1 W 1 W 2 If D 1 or D 2 has a finite optimum value, then W 1 = W 2
11 Propositions Duality Theorem Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example If G = L r or Γ r, problem D 2 is a strong dual of problem P 1. If P has a finite optimum value, there is an optimal dual function of the form: F(d) = P r i=1 u id i + F 2 (d r+1,..., d m), u i 0, i = 1,..., r, F 2 Γ m r if and only if Z = W nx 1 = min max{ (c j uh j )x j + uh : Ex e, x 0, integer} u 0 j=1 if and only if max{cx : Hx h, x conv[ex e, x 0, integral]} has an optimal solution in integers. If P has a finite optimum value, it has an optimal dual price function that is separable, F Γ r Γ m r if and only if max{cx : x conv[hx h, x 0, integral] x conv[ex e, x 0, integral]} has an optimal solution in integers.
12 Gomory fractional cutting plane algorithm Initialization: Set r = 1. Iteration r: Solve the problem P r 1 : Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example Z r 1 = max s.t nx c j x j j=1 nx a ij x j b i i = 1,..., m + r 1 j=1 x 0 If P r 1 is infeasible or the solution is integer, stop. Add the cut P n j=1 a (m+r)jx j b m+r which has the form: nx mx f r (a j )x j f r (b) where f r (d) = j=1 i=1 Xr 1 i d i + λ r 1 m+i f i (d) and λ r 1 i=1 λ r+i = (λ r 1 1,..., λ r 1 m+r 1 ) 0
13 Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example If the Gomory cutting plane algorithm terminates finitely and P has a finite optimum value, there exists an optimal value function of the form F(d) = G r (d) = mx rx ui r d i + um+if r i (d) i=1 i=1 for some finite value r and if u r is dual optimal, Z r = G r (b)
14 Branch and Bound Algorithm Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example Replace P by a family of problems: x is optimal if: (P t) Z t = max{cx : Ax, x X t}, t = 1,..., r where x is optimal for some problem P t r t=1x t = {x : x 0, integer} Z t = cx Z t t t where Z t is an upper bound on the optimal value of P t In this case it is not true that the algorithms generate an optimal superadditive value function. However, an optimal dual function with some other structure can be constructed.
15 Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example It is shown that the problem min F(b) s.t F(Ax) cx x X t F {F : R m R {± }} and nondecreasing is a strong dual of problem P t. Hence, the problem min F(b) s.t F(Ax) cx x 0, integer ( D) F {F : R m R} and nondecreasing is a strong dual of problem P. Let X t = {x : 0 g t j x j h t j, j = 1,..., n, x 0, integer}. Then if a standard branch and bound algorithm terminates with a finite optimal value, then D has an optimal solution of the form: for some finite value of r. F (d) = max t=1,...,r [αt + π t d], π t 0
16 Alternative formulation Duality Theorem Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example Let P be the problem: Z = max{cx : Ax b, x g, x h, x 0, integer} with the same X t and r t=1x t = {x : 0 g x h, x 0, integer}. P has a dual problem D: D = min{f(b, g, h) : F(a j, e j, e j ) c j, j = 1,..., n, F Γ m+2n } If a standard branch and bound algorithm terminates with a finite optimal value, then D has a feasible solution of the form: F(d) = min t=1,...,r ut d and nx j=1 F(a j, e j, e j )x j max t=1,...,r F(b, gt, h t ) (1) is a valid cut for P.
17 Separability & Lagrangean Relaxation Cutting Planes Branch and Bound Example
18 Duality Theorem Characterization Solution Procedure For the problem min cx s.t Ax = b, A Z m + n, b Z m +, c Z n x Z n + given a vector α R m, we define a generator subadditive function F α : R m + R as P F α(d) = αd max i E (αa i c i )x i A E x d x Z E + where E = {i N : αa i > c i } and A E is the submatrix of A consisting of the columns with indices in E. Also we define a divergent generator subadditive function F β : R m + R as F β (d) = βd max P i E (βa i)x i A E x d x Z E +
19 Characterization Solution Procedure b = ˆ c = ˆ A =
20 Properties Duality Theorem Characterization Solution Procedure Weak duality: F α is feasible subadditive function and hence F α(b) cx, α and feasible primal feasible vectors. Strong duality: If the IP is feasible, then there exists an α s.t F α is a generator optimal subadditive function(ofs),i.e., F α(b) = z IP. If the IP is infeasible, then there exists a divergent generator subadditive function F β s.t Fβ > 0 Example Complementary slackness: Let x be an optimal IP solution. If x i > 0 then αa i c i in any generator OSF.
21 Properties Duality Theorem Characterization Solution Procedure Reduced cost fixing: Let F be a feasible subadditive dual function and let z IP be an upper bound on z IP. If c k F(a k ) > 0 and v = zip F(b) c k F(a k ) > 0 for a column k N, then there is an optimal IP solution x with x k v 1 Valid Inequalities: For any generator feasible subadditive function X c i x i + X (αa i )x i F α(b) (2) i E i H is a valid inequality. Example
22 Basic generator subadditive functions Characterization Solution Procedure For an E N we can find a generator subadditive function F α with the best objective value and such that E(α ) E by the problem P E max{η : (η, α) Q b (E)} s.t Q b (E) = {η + α(a E x b) c E x x Z E +, A E x b (η, α) R R m } αa i c i i H = N \ E A generator subadditive function F α is called a basic generator subadditive function(bg) if (F α(b), α) is an extreme point of the polyhedron Q b (E(α)). (Similar argument for divergent generator subadditive functions). BG functions yield facet-defining valid inequalities and hence suffice to solve the IP as an LP.
23 Main frame Duality Theorem Characterization Solution Procedure Given E solve P E to get an α. Next set E = E i and update α. Iterate until F α(b) = cx. However P E is hard to solve. Therefore, use row generation. Let Ê Zn + s.t. Ê {x Zn + : A E x b} where E = x Ê {i N : x i > 0} Note that if A E x b, then x is not necessarily in Ê. Hence, subadditivity may not be satisfied.
24 Maintaining subadditivity Duality Theorem Characterization Solution Procedure Let S(x) = {y Z n + : y x}. Given Ê Zn + and a vector α, define π(x) = αax max {(αa c)y} y Ê,y S(x) π is dual feasible if π(e i ) c i, i N. If π is dual feasible and subadditive and x is feasible to the IP, then π(x) X i N π(e i )x i cx Hence, π provides weak duality. In addition consider Ê = {x Z E +, A E x b} where E is defined based on the α value of an OSF. Then, strong duality follows.
25 D(Ê, Ĥ) Duality Theorem Characterization Solution Procedure We have the new problem: max π max π(x) x Z n +,Ax=b π(e i ) c i i N π subadditive Alternatively, α that gives the largest dual objective value can be obtained from D(Ê, Ĥ): max π 0 π 0 + α(ay b) cy y Ê αax cx x Ĥ where Ĥ = {x Zn + : x / Ê, S(x) \ {x} Ê}.
26 The Algorithm Duality Theorem Characterization Solution Procedure Ê =, Ĥ = {e i : i N}, α = optimal dual vector of the LP relaxation, w D = LOOP Choose a vector x Ĥ, Ê = Ê { x}, Ĥ = Ĥ { x + e i : i N, y Ê, y x + e i, y Z n +} Update α by solving D(Ê, Ĥ). Let π 0 be the optimal value. w D = max{w D, π0 } If w D = min{cx : x Ĥ, Ax = b}, then we have solved the IP and exit. END LOOP
27 Computing a generator OSF Characterization Solution Procedure It has 3 stages: Find π based on the previous algorithm. Improve π and get it subadditive in {x Z S + : A S x = b, c s x z IP } for an S N. Obtain a generator OSF. Start with E = x Ê {i N : x i > 0} where Ê is obtained in previous stage. Keep expanding E and solve P E till getting the optimal OSF.
28 With an optimal solution pair (x, F ) With an OSF With Branch and Bound Inference Duality Changes in the rhs: b b F remains feasible. Z F (b ) x {y F ( P n j=1 a j y j ) = P n j=1 c j y j } if F is still optimal. Changes in the objective: c c x remains primal feasible. Z c x If c j F (a j ) j, then F remains dual feasible. Z F (b) If c j F (a j ) when xj = 0 and c j = c j when xj > 0, then x remains optimal.
29 With an OSF With Branch and Bound Inference Duality Adding a new column: ( c, ã) (x, 0) remains primal feasible. Z Z (x, 0) remains optimal if F (ã) c Adding a new constraint: P n j=1 h jx j h 0 If x is still feasible, x is optimal. The price function F : R m+1 R defined by F(d, d m+1 ) = F (d) is dual feasible for the new problem. Z Z = F(b, h 0 )
30 After a mixed-binary program is solved With an OSF With Branch and Bound Inference Duality Branch and bound algorithm exits with a tree of LP problems of the form (for node t): Q t (b) : Z t (b) = max cx nx s.t. a j x j b (π t ) where j=1 x j 0 for j N (θ t j ) L t j x j U t j for j I (µ t j ) L t j = Uj t = 0 for j F0 t L t j = Uj t = 1 for j F1 t L t j = 0, U t j = 1 for j I F t 1 F t 0 and π t, µ t are the dual feasible prices, I = {j N : x j {0, 1}} and F t 0, F t 1 are the sets of variables in I fixed to 0 and 1 respectively.
31 With an OSF With Branch and Bound Inference Duality Let C t (d) and B t (d) be the upper bounding functions on Q t (d) and the corresponding mixed-ip problem. We want to get B 1 (d). For each node t: C t (d) = π t d + X j I F t 0 F t 1 max{0, µ t j } + X j F t 1 For each terminal node with a feasible solution to Q t (b): B t (d) = C t (d) For a terminal node with no feasible solution: B t (d) = { if Ct (d) < 0 + if C t (d) 0 For any noterminal node having two offsprings L(t) and R(t): B t (d) = min {C t (d), max {B L(t) (d), B R(t) (d)}} µ t j
32 With an OSF With Branch and Bound Inference Duality Z B 1 (d) when problem is solved with a rhs of d. If P has been solved to optimality and a new column a j is added to the problem, then there exists an optimal solution to the new problem with x j = 0 if c j B 1 (b) B 1 (b a j )
33 With an OSF With Branch and Bound Inference Duality Next time...
34 Methods to Solve PIPs Parametric objective MIP z(φ) = min (c + φc )x s.t Ax b (Pφ) x 0 x j integer, j J N Parametric rhs MIP z(θ) = min cx s.t Ax (b + θb ) (Pθ) x 0 x j integer, j J N Parametric A-matrix MIP z(λ) = min cx s.t (A + λa )x b (Pλ) x 0 x j integer, j J N where φ, θ, α [0, 1] and c, b, A are the change vectors and the matrix.
35 Methods to Solve PIPs By solving Pφ for instance, we mean to determine an optimal solution and its value, x (φ) and z (φ) for all 0 φ 1. z(φ) and z(θ) - Graphs here The solution procedure usually follows to solve a number of problems Pφ k, k = 1,..K with φ k determined through an algorithm.
36 Objective case - IP Duality Theorem Methods to Solve PIPs Concavity may be used to define upper and lower bounds on z(φ) : z LB (φ), z UB (φ) Graph Some φ is removed from the set {φ i } and P φ is solved. If z( φ) = z UB ( φ) then z(φ) = z UB (φ) for all φ ˆ φl, φ R. If z( φ) z UB ( φ), then a new x( φ) has been found. Then the intersection point of (c + φc )x( φ) with the parts UB( φ) define new values φ i and φ i to the left and the right of φ respectively. Jenkins showed that this procedure requires solving 2p-1 point-values IPs where p is the number of different optimal solutions found for φ [0, 1]. He uses cutting planes algorithm together with some post optimality analysis to determine ranges for φ.
37 RHS Case - IP Duality Theorem Methods to Solve PIPs The hyperplane defined by constraint i cannot pass any integral point unless the integer portion of b i + θb i changes. Let d = 1 lcm{b 1,..., b n} Then, any b i + θb i, for i = 1,..., m will never be integer when (w 1)d < θ < wd with w = 1,..., 1/d. Thus, all of the m hyperplanes will not pass any integral point during these intervals of θ. Now, let wd with w = 0, 1,...1/d be the candidates of θ.
38 Methods to Solve PIPs When θ varies, the current optimal solution will not change until at least one constraint passes some integral point not every candidate of θ will make b i + θb i integer such that the i th constraint is binding at some integral point. Let these θ values be the principal candidates. If ζ r and ζ r+1 are two adjacent principal candidates of θ, then when θ varies within (ζ r, ζ r+1 ), there exists at most one optimal solution to Pθ If ζ r and ζ r+1 are two adjacent principal candidates and x(ζ r ) = x(ζ r+1 ), then x(ζ r ) is optimal for all ζ r θ ζ r+1 Using some elimination techniques(tolerance intervals for the optimal solutions), solve a sequence of candidate problems to get Pθ
39 Methods to Solve PIPs A branch and cut algorithm for general monotone (b 0) parametric IP It solves R -the number of principal candidates- problems simulatenously, i.e., min cx s.t Ax (b + t r) (P tr ) x 0 integer where t r = θ r b.
40 Methods to Solve PIPs Use dual simplex method on the LP-relaxations of all R problems till the relaxation of P b is at optimality. Save results as LBs on members. Add cuts to the first primal feasible and unfathomed member and resolve Check for integer solutions and update UBs. Note that UB(P t) is valid for another member with a smaller t and the members are fathomed from a node if UB(P t) LB(P t). If a member is fathomed, than move to the next member in decreasing order. If it is not primal feasible, pivot until optimality. Repeat this step. Branch. Select a node and pivot on the first unfathomed member. Repeat the previous step until all nodes are fathomed. Heuristics can also be used to improve the UBs.
41 Methods to Solve PIPs A branch and bound algorithm for binary IP A standart branch and bound algorithm with upper and lower bounds depending on θ. UB(θ): For the known feasible solutions x k, k K define: θ k 1 = min{θ θ k 2 = max{θ nx j=1 nx j=1 A j x k j b + θb } A j x k j b + θb } where θ k 1 = θ k 2 = if the indicated set is empty. Then Hence, UB k (θ) = { nx j=1 c j x k j if θ k 1 θ θ k 2, otherwise} UB(θ) = min {UB k (θ) k K }
42 Methods to Solve PIPs LB q (θ) of a node q is basically the LP relaxation of P q θ. A lower bound function to LB q (θ) can be derived from an optimal dual solution. If the primal problem is feasible: LB q (θ) (u q b )θ + u q b + v q e where u q and v q are the optimal dual values. If the primal problem is infeasible, then there exists an extreme ray (y q, z q ) 0 such that y q b + z q e > 0. If y q b 0, then LB q (θ) = + for all 0 θ 1. Otherwise, LB q (θ) = for all 0 θ θ. and LB q (θ) (u q b )θ + u q b + v q e for all θ θ 1 where (u q, v q ) is an extreme point and Graph of LB q (θ) θ = y q (b) z q e y q b
43 Methods to Solve PIPs Begin with P 0. Update UB(θ) when an integer solution is obtained. Besides, use heuristics for several values of theta to get integer solutions. Fathoming: if LB q (θ) UB(θ) for all 0 θ 1, then the corresponding node will not yield a better descendent. (Do not fathom even the optimal primal solution is all integer). When the algorithm terminates, P(θ) = UB(θ) Example
44 Methods to Solve PIPs
45 Next Time Duality Theorem Methods to Solve PIPs A subset of More of IP Duality MIP Duality Sensitivity Analysis, Inference based SA Parametric Integer Programming Warm Starting
A New Subadditive Approach to Integer Programming: Theory and Algorithms
A New Subadditive Approach to Integer Programming: Theory and Algorithms Diego Klabjan Department of Mechanical and Industrial Engineering University of Illinois at Urbana-Champaign Urbana, IL email: klabjan@uiuc.edu
More informationSection Notes 9. IP: Cutting Planes. Applied Math 121. Week of April 12, 2010
Section Notes 9 IP: Cutting Planes Applied Math 121 Week of April 12, 2010 Goals for the week understand what a strong formulations is. be familiar with the cutting planes algorithm and the types of cuts
More informationSection Notes 9. Midterm 2 Review. Applied Math / Engineering Sciences 121. Week of December 3, 2018
Section Notes 9 Midterm 2 Review Applied Math / Engineering Sciences 121 Week of December 3, 2018 The following list of topics is an overview of the material that was covered in the lectures and sections
More informationInteger Programming Duality
Integer Programming Duality M. Guzelsoy T. K. Ralphs July, 2010 1 Introduction This article describes what is known about duality for integer programs. It is perhaps surprising that many of the results
More informationIntroduction to Integer Linear Programming
Lecture 7/12/2006 p. 1/30 Introduction to Integer Linear Programming Leo Liberti, Ruslan Sadykov LIX, École Polytechnique liberti@lix.polytechnique.fr sadykov@lix.polytechnique.fr Lecture 7/12/2006 p.
More informationDuality, Warm Starting, and Sensitivity Analysis for MILP
Duality, Warm Starting, and Sensitivity Analysis for MILP Ted Ralphs and Menal Guzelsoy Industrial and Systems Engineering Lehigh University SAS Institute, Cary, NC, Friday, August 19, 2005 SAS Institute
More informationNetwork Flows. 6. Lagrangian Relaxation. Programming. Fall 2010 Instructor: Dr. Masoud Yaghini
In the name of God Network Flows 6. Lagrangian Relaxation 6.3 Lagrangian Relaxation and Integer Programming Fall 2010 Instructor: Dr. Masoud Yaghini Integer Programming Outline Branch-and-Bound Technique
More informationInteger Programming ISE 418. Lecture 8. Dr. Ted Ralphs
Integer Programming ISE 418 Lecture 8 Dr. Ted Ralphs ISE 418 Lecture 8 1 Reading for This Lecture Wolsey Chapter 2 Nemhauser and Wolsey Sections II.3.1, II.3.6, II.4.1, II.4.2, II.5.4 Duality for Mixed-Integer
More informationIntroduction to integer programming II
Introduction to integer programming II Martin Branda Charles University in Prague Faculty of Mathematics and Physics Department of Probability and Mathematical Statistics Computational Aspects of Optimization
More informationSection Notes 8. Integer Programming II. Applied Math 121. Week of April 5, expand your knowledge of big M s and logical constraints.
Section Notes 8 Integer Programming II Applied Math 121 Week of April 5, 2010 Goals for the week understand IP relaxations be able to determine the relative strength of formulations understand the branch
More information3.7 Cutting plane methods
3.7 Cutting plane methods Generic ILP problem min{ c t x : x X = {x Z n + : Ax b} } with m n matrix A and n 1 vector b of rationals. According to Meyer s theorem: There exists an ideal formulation: conv(x
More informationCutting Plane Separators in SCIP
Cutting Plane Separators in SCIP Kati Wolter Zuse Institute Berlin DFG Research Center MATHEON Mathematics for key technologies 1 / 36 General Cutting Plane Method MIP min{c T x : x X MIP }, X MIP := {x
More informationmaxz = 3x 1 +4x 2 2x 1 +x 2 6 2x 1 +3x 2 9 x 1,x 2
ex-5.-5. Foundations of Operations Research Prof. E. Amaldi 5. Branch-and-Bound Given the integer linear program maxz = x +x x +x 6 x +x 9 x,x integer solve it via the Branch-and-Bound method (solving
More informationMVE165/MMG631 Linear and integer optimization with applications Lecture 8 Discrete optimization: theory and algorithms
MVE165/MMG631 Linear and integer optimization with applications Lecture 8 Discrete optimization: theory and algorithms Ann-Brith Strömberg 2017 04 07 Lecture 8 Linear and integer optimization with applications
More informationCutting Planes in SCIP
Cutting Planes in SCIP Kati Wolter Zuse-Institute Berlin Department Optimization Berlin, 6th June 2007 Outline 1 Cutting Planes in SCIP 2 Cutting Planes for the 0-1 Knapsack Problem 2.1 Cover Cuts 2.2
More information3.10 Lagrangian relaxation
3.10 Lagrangian relaxation Consider a generic ILP problem min {c t x : Ax b, Dx d, x Z n } with integer coefficients. Suppose Dx d are the complicating constraints. Often the linear relaxation and the
More informationAM 121: Intro to Optimization! Models and Methods! Fall 2018!
AM 121: Intro to Optimization Models and Methods Fall 2018 Lecture 15: Cutting plane methods Yiling Chen SEAS Lesson Plan Cut generation and the separation problem Cutting plane methods Chvatal-Gomory
More informationLagrangian Relaxation in MIP
Lagrangian Relaxation in MIP Bernard Gendron May 28, 2016 Master Class on Decomposition, CPAIOR2016, Banff, Canada CIRRELT and Département d informatique et de recherche opérationnelle, Université de Montréal,
More informationColumn Generation. MTech Seminar Report. Soumitra Pal Roll No: under the guidance of
Column Generation MTech Seminar Report by Soumitra Pal Roll No: 05305015 under the guidance of Prof. A. G. Ranade Computer Science and Engineering IIT-Bombay a Department of Computer Science and Engineering
More informationA Review of Linear Programming
A Review of Linear Programming Instructor: Farid Alizadeh IEOR 4600y Spring 2001 February 14, 2001 1 Overview In this note we review the basic properties of linear programming including the primal simplex
More informationReview Solutions, Exam 2, Operations Research
Review Solutions, Exam 2, Operations Research 1. Prove the weak duality theorem: For any x feasible for the primal and y feasible for the dual, then... HINT: Consider the quantity y T Ax. SOLUTION: To
More informationConvex Analysis 2013 Let f : Q R be a strongly convex function with convexity parameter µ>0, where Q R n is a bounded, closed, convex set, which contains the origin. Let Q =conv(q, Q) andconsiderthefunction
More informationOn the Value Function of a Mixed Integer Linear Program
On the Value Function of a Mixed Integer Linear Program MENAL GUZELSOY TED RALPHS ISE Department COR@L Lab Lehigh University ted@lehigh.edu AIRO, Ischia, Italy, 11 September 008 Thanks: Work supported
More informationYinyu Ye, MS&E, Stanford MS&E310 Lecture Note #06. The Simplex Method
The Simplex Method Yinyu Ye Department of Management Science and Engineering Stanford University Stanford, CA 94305, U.S.A. http://www.stanford.edu/ yyye (LY, Chapters 2.3-2.5, 3.1-3.4) 1 Geometry of Linear
More informationChapter 1. Preliminaries
Introduction This dissertation is a reading of chapter 4 in part I of the book : Integer and Combinatorial Optimization by George L. Nemhauser & Laurence A. Wolsey. The chapter elaborates links between
More informationInteger Programming. Wolfram Wiesemann. December 6, 2007
Integer Programming Wolfram Wiesemann December 6, 2007 Contents of this Lecture Revision: Mixed Integer Programming Problems Branch & Bound Algorithms: The Big Picture Solving MIP s: Complete Enumeration
More information3.8 Strong valid inequalities
3.8 Strong valid inequalities By studying the problem structure, we can derive strong valid inequalities which lead to better approximations of the ideal formulation conv(x ) and hence to tighter bounds.
More informationmin3x 1 + 4x 2 + 5x 3 2x 1 + 2x 2 + x 3 6 x 1 + 2x 2 + 3x 3 5 x 1, x 2, x 3 0.
ex-.-. Foundations of Operations Research Prof. E. Amaldi. Dual simplex algorithm Given the linear program minx + x + x x + x + x 6 x + x + x x, x, x. solve it via the dual simplex algorithm. Describe
More informationInteger Programming ISE 418. Lecture 13. Dr. Ted Ralphs
Integer Programming ISE 418 Lecture 13 Dr. Ted Ralphs ISE 418 Lecture 13 1 Reading for This Lecture Nemhauser and Wolsey Sections II.1.1-II.1.3, II.1.6 Wolsey Chapter 8 CCZ Chapters 5 and 6 Valid Inequalities
More information- Well-characterized problems, min-max relations, approximate certificates. - LP problems in the standard form, primal and dual linear programs
LP-Duality ( Approximation Algorithms by V. Vazirani, Chapter 12) - Well-characterized problems, min-max relations, approximate certificates - LP problems in the standard form, primal and dual linear programs
More information(P ) Minimize 4x 1 + 6x 2 + 5x 3 s.t. 2x 1 3x 3 3 3x 2 2x 3 6
The exam is three hours long and consists of 4 exercises. The exam is graded on a scale 0-25 points, and the points assigned to each question are indicated in parenthesis within the text. Problem 1 Consider
More informationDiscrete (and Continuous) Optimization WI4 131
Discrete (and Continuous) Optimization WI4 131 Kees Roos Technische Universiteit Delft Faculteit Electrotechniek, Wiskunde en Informatica Afdeling Informatie, Systemen en Algoritmiek e-mail: C.Roos@ewi.tudelft.nl
More informationMath Models of OR: Branch-and-Bound
Math Models of OR: Branch-and-Bound John E. Mitchell Department of Mathematical Sciences RPI, Troy, NY 12180 USA November 2018 Mitchell Branch-and-Bound 1 / 15 Branch-and-Bound Outline 1 Branch-and-Bound
More informationSpring 2017 CO 250 Course Notes TABLE OF CONTENTS. richardwu.ca. CO 250 Course Notes. Introduction to Optimization
Spring 2017 CO 250 Course Notes TABLE OF CONTENTS richardwu.ca CO 250 Course Notes Introduction to Optimization Kanstantsin Pashkovich Spring 2017 University of Waterloo Last Revision: March 4, 2018 Table
More informationmin 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
The exam is three hours long and consists of 4 exercises. The exam is graded on a scale 0-25 points, and the points assigned to each question are indicated in parenthesis within the text. If necessary,
More informationReformulation and Decomposition of Integer Programs
Reformulation and Decomposition of Integer Programs François Vanderbeck 1 and Laurence A. Wolsey 2 (Reference: CORE DP 2009/16) (1) Université Bordeaux 1 & INRIA-Bordeaux (2) Université de Louvain, CORE.
More informationSolving LP and MIP Models with Piecewise Linear Objective Functions
Solving LP and MIP Models with Piecewise Linear Obective Functions Zonghao Gu Gurobi Optimization Inc. Columbus, July 23, 2014 Overview } Introduction } Piecewise linear (PWL) function Convex and convex
More informationLecture: Algorithms for LP, SOCP and SDP
1/53 Lecture: Algorithms for LP, SOCP and SDP Zaiwen Wen Beijing International Center For Mathematical Research Peking University http://bicmr.pku.edu.cn/~wenzw/bigdata2018.html wenzw@pku.edu.cn Acknowledgement:
More informationAppendix PRELIMINARIES 1. THEOREMS OF ALTERNATIVES FOR SYSTEMS OF LINEAR CONSTRAINTS
Appendix PRELIMINARIES 1. THEOREMS OF ALTERNATIVES FOR SYSTEMS OF LINEAR CONSTRAINTS Here we consider systems of linear constraints, consisting of equations or inequalities or both. A feasible solution
More informationLP Duality: outline. Duality theory for Linear Programming. alternatives. optimization I Idea: polyhedra
LP Duality: outline I Motivation and definition of a dual LP I Weak duality I Separating hyperplane theorem and theorems of the alternatives I Strong duality and complementary slackness I Using duality
More informationAn Integer Cutting-Plane Procedure for the Dantzig-Wolfe Decomposition: Theory
An Integer Cutting-Plane Procedure for the Dantzig-Wolfe Decomposition: Theory by Troels Martin Range Discussion Papers on Business and Economics No. 10/2006 FURTHER INFORMATION Department of Business
More informationLinear and Integer Programming - ideas
Linear and Integer Programming - ideas Paweł Zieliński Institute of Mathematics and Computer Science, Wrocław University of Technology, Poland http://www.im.pwr.wroc.pl/ pziel/ Toulouse, France 2012 Literature
More informationInteger Programming ISE 418. Lecture 12. Dr. Ted Ralphs
Integer Programming ISE 418 Lecture 12 Dr. Ted Ralphs ISE 418 Lecture 12 1 Reading for This Lecture Nemhauser and Wolsey Sections II.2.1 Wolsey Chapter 9 ISE 418 Lecture 12 2 Generating Stronger Valid
More informationDiscrete Optimization 2010 Lecture 7 Introduction to Integer Programming
Discrete Optimization 2010 Lecture 7 Introduction to Integer Programming Marc Uetz University of Twente m.uetz@utwente.nl Lecture 8: sheet 1 / 32 Marc Uetz Discrete Optimization Outline 1 Intro: The Matching
More informationBBM402-Lecture 20: LP Duality
BBM402-Lecture 20: LP Duality Lecturer: Lale Özkahya Resources for the presentation: https://courses.engr.illinois.edu/cs473/fa2016/lectures.html An easy LP? which is compact form for max cx subject to
More informationSOLVING INTEGER LINEAR PROGRAMS. 1. Solving the LP relaxation. 2. How to deal with fractional solutions?
SOLVING INTEGER LINEAR PROGRAMS 1. Solving the LP relaxation. 2. How to deal with fractional solutions? Integer Linear Program: Example max x 1 2x 2 0.5x 3 0.2x 4 x 5 +0.6x 6 s.t. x 1 +2x 2 1 x 1 + x 2
More informationInterior-Point versus Simplex methods for Integer Programming Branch-and-Bound
Interior-Point versus Simplex methods for Integer Programming Branch-and-Bound Samir Elhedhli elhedhli@uwaterloo.ca Department of Management Sciences, University of Waterloo, Canada Page of 4 McMaster
More information23. Cutting planes and branch & bound
CS/ECE/ISyE 524 Introduction to Optimization Spring 207 8 23. Cutting planes and branch & bound ˆ Algorithms for solving MIPs ˆ Cutting plane methods ˆ Branch and bound methods Laurent Lessard (www.laurentlessard.com)
More information9.1 Linear Programs in canonical form
9.1 Linear Programs in canonical form LP in standard form: max (LP) s.t. where b i R, i = 1,..., m z = j c jx j j a ijx j b i i = 1,..., m x j 0 j = 1,..., n But the Simplex method works only on systems
More informationLinear Programming Inverse Projection Theory Chapter 3
1 Linear Programming Inverse Projection Theory Chapter 3 University of Chicago Booth School of Business Kipp Martin September 26, 2017 2 Where We Are Headed We want to solve problems with special structure!
More informationLagrangean relaxation
Lagrangean relaxation Giovanni Righini Corso di Complementi di Ricerca Operativa Joseph Louis de la Grange (Torino 1736 - Paris 1813) Relaxations Given a problem P, such as: minimize z P (x) s.t. x X P
More informationPart 1. The Review of Linear Programming
In the name of God Part 1. The Review of Linear Programming 1.5. Spring 2010 Instructor: Dr. Masoud Yaghini Outline Introduction Formulation of the Dual Problem Primal-Dual Relationship Economic Interpretation
More informationThe Primal-Dual Algorithm P&S Chapter 5 Last Revised October 30, 2006
The Primal-Dual Algorithm P&S Chapter 5 Last Revised October 30, 2006 1 Simplex solves LP by starting at a Basic Feasible Solution (BFS) and moving from BFS to BFS, always improving the objective function,
More informationThe Value function of a Mixed-Integer Linear Program with a Single Constraint
The Value Function of a Mixed Integer Linear Programs with a Single Constraint MENAL GUZELSOY TED RALPHS ISE Department COR@L Lab Lehigh University tkralphs@lehigh.edu University of Wisconsin February
More informationto work with) can be solved by solving their LP relaxations with the Simplex method I Cutting plane algorithms, e.g., Gomory s fractional cutting
Summary so far z =max{c T x : Ax apple b, x 2 Z n +} I Modeling with IP (and MIP, and BIP) problems I Formulation for a discrete set that is a feasible region of an IP I Alternative formulations for the
More informationLessons from MIP Search. John Hooker Carnegie Mellon University November 2009
Lessons from MIP Search John Hooker Carnegie Mellon University November 2009 Outline MIP search The main ideas Duality and nogoods From MIP to AI (and back) Binary decision diagrams From MIP to constraint
More informationSlack Variable. Max Z= 3x 1 + 4x 2 + 5X 3. Subject to: X 1 + X 2 + X x 1 + 4x 2 + X X 1 + X 2 + 4X 3 10 X 1 0, X 2 0, X 3 0
Simplex Method Slack Variable Max Z= 3x 1 + 4x 2 + 5X 3 Subject to: X 1 + X 2 + X 3 20 3x 1 + 4x 2 + X 3 15 2X 1 + X 2 + 4X 3 10 X 1 0, X 2 0, X 3 0 Standard Form Max Z= 3x 1 +4x 2 +5X 3 + 0S 1 + 0S 2
More informationStructured Problems and Algorithms
Integer and quadratic optimization problems Dept. of Engg. and Comp. Sci., Univ. of Cal., Davis Aug. 13, 2010 Table of contents Outline 1 2 3 Benefits of Structured Problems Optimization problems may become
More informationA notion of Total Dual Integrality for Convex, Semidefinite and Extended Formulations
A notion of for Convex, Semidefinite and Extended Formulations Marcel de Carli Silva Levent Tunçel April 26, 2018 A vector in R n is integral if each of its components is an integer, A vector in R n is
More information4. Duality and Sensitivity
4. Duality and Sensitivity For every instance of an LP, there is an associated LP known as the dual problem. The original problem is known as the primal problem. There are two de nitions of the dual pair
More informationOn the Polyhedral Structure of a Multi Item Production Planning Model with Setup Times
CORE DISCUSSION PAPER 2000/52 On the Polyhedral Structure of a Multi Item Production Planning Model with Setup Times Andrew J. Miller 1, George L. Nemhauser 2, and Martin W.P. Savelsbergh 2 November 2000
More informationChapter 1 Linear Programming. Paragraph 5 Duality
Chapter 1 Linear Programming Paragraph 5 Duality What we did so far We developed the 2-Phase Simplex Algorithm: Hop (reasonably) from basic solution (bs) to bs until you find a basic feasible solution
More informationIP Cut Homework from J and B Chapter 9: 14, 15, 16, 23, 24, You wish to solve the IP below with a cutting plane technique.
IP Cut Homework from J and B Chapter 9: 14, 15, 16, 23, 24, 31 14. You wish to solve the IP below with a cutting plane technique. Maximize 4x 1 + 2x 2 + x 3 subject to 14x 1 + 10x 2 + 11x 3 32 10x 1 +
More information3.7 Strong valid inequalities for structured ILP problems
3.7 Strong valid inequalities for structured ILP problems By studying the problem structure, we can derive strong valid inequalities yielding better approximations of conv(x ) and hence tighter bounds.
More informationSubadditive Approaches to Mixed Integer Programming
Subadditive Approaches to Mixed Integer Programming by Babak Moazzez A thesis submitted to the Faculty of Graduate and Postdoctoral Affairs in partial fulfillment of the requirements for the degree of
More informationSummary of the simplex method
MVE165/MMG630, The simplex method; degeneracy; unbounded solutions; infeasibility; starting solutions; duality; interpretation Ann-Brith Strömberg 2012 03 16 Summary of the simplex method Optimality condition:
More information1 Review Session. 1.1 Lecture 2
1 Review Session Note: The following lists give an overview of the material that was covered in the lectures and sections. Your TF will go through these lists. If anything is unclear or you have questions
More information21. Solve the LP given in Exercise 19 using the big-m method discussed in Exercise 20.
Extra Problems for Chapter 3. Linear Programming Methods 20. (Big-M Method) An alternative to the two-phase method of finding an initial basic feasible solution by minimizing the sum of the artificial
More informationAdvanced Linear Programming: The Exercises
Advanced Linear Programming: The Exercises The answers are sometimes not written out completely. 1.5 a) min c T x + d T y Ax + By b y = x (1) First reformulation, using z smallest number satisfying x z
More information"SYMMETRIC" PRIMAL-DUAL PAIR
"SYMMETRIC" PRIMAL-DUAL PAIR PRIMAL Minimize cx DUAL Maximize y T b st Ax b st A T y c T x y Here c 1 n, x n 1, b m 1, A m n, y m 1, WITH THE PRIMAL IN STANDARD FORM... Minimize cx Maximize y T b st Ax
More informationDecomposition and Reformulation in Integer Programming
and Reformulation in Integer Programming Laurence A. WOLSEY 7/1/2008 / Aussois and Reformulation in Integer Programming Outline 1 Resource 2 and Reformulation in Integer Programming Outline Resource 1
More informationMVE165/MMG631 Linear and integer optimization with applications Lecture 5 Linear programming duality and sensitivity analysis
MVE165/MMG631 Linear and integer optimization with applications Lecture 5 Linear programming duality and sensitivity analysis Ann-Brith Strömberg 2017 03 29 Lecture 4 Linear and integer optimization with
More informationI.3. LMI DUALITY. Didier HENRION EECI Graduate School on Control Supélec - Spring 2010
I.3. LMI DUALITY Didier HENRION henrion@laas.fr EECI Graduate School on Control Supélec - Spring 2010 Primal and dual For primal problem p = inf x g 0 (x) s.t. g i (x) 0 define Lagrangian L(x, z) = g 0
More informationIntroduction to Mathematical Programming IE406. Lecture 10. Dr. Ted Ralphs
Introduction to Mathematical Programming IE406 Lecture 10 Dr. Ted Ralphs IE406 Lecture 10 1 Reading for This Lecture Bertsimas 4.1-4.3 IE406 Lecture 10 2 Duality Theory: Motivation Consider the following
More informationIn the original knapsack problem, the value of the contents of the knapsack is maximized subject to a single capacity constraint, for example weight.
In the original knapsack problem, the value of the contents of the knapsack is maximized subject to a single capacity constraint, for example weight. In the multi-dimensional knapsack problem, additional
More informationBilevel Integer Linear Programming
Bilevel Integer Linear Programming TED RALPHS SCOTT DENEGRE ISE Department COR@L Lab Lehigh University ted@lehigh.edu MOPTA 2009, Lehigh University, 19 August 2009 Thanks: Work supported in part by the
More informationLift-and-Project Inequalities
Lift-and-Project Inequalities Q. Louveaux Abstract The lift-and-project technique is a systematic way to generate valid inequalities for a mixed binary program. The technique is interesting both on the
More informationNote 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)
Note 3: LP Duality If the primal problem (P) in the canonical form is min Z = n j=1 c j x j s.t. nj=1 a ij x j b i i = 1, 2,..., m (1) x j 0 j = 1, 2,..., n, then the dual problem (D) in the canonical
More informationOptimization WS 13/14:, by Y. Goldstein/K. Reinert, 9. Dezember 2013, 16: Linear programming. Optimization Problems
Optimization WS 13/14:, by Y. Goldstein/K. Reinert, 9. Dezember 2013, 16:38 2001 Linear programming Optimization Problems General optimization problem max{z(x) f j (x) 0,x D} or min{z(x) f j (x) 0,x D}
More informationF 1 F 2 Daily Requirement Cost N N N
Chapter 5 DUALITY 5. The Dual Problems Every linear programming problem has associated with it another linear programming problem and that the two problems have such a close relationship that whenever
More informationSimplex Algorithm Using Canonical Tableaus
41 Simplex Algorithm Using Canonical Tableaus Consider LP in standard form: Min z = cx + α subject to Ax = b where A m n has rank m and α is a constant In tableau form we record it as below Original Tableau
More informationOverview of course. Introduction to Optimization, DIKU Monday 12 November David Pisinger
Introduction to Optimization, DIKU 007-08 Monday November David Pisinger Lecture What is OR, linear models, standard form, slack form, simplex repetition, graphical interpretation, extreme points, basic
More informationDuality for Mixed-Integer Linear Programs
Duality for Mixed-Integer Linear Programs M. Guzelsoy T. K. Ralphs Original May, 006 Revised April, 007 Abstract The theory of duality for linear programs is well-developed and has been successful in advancing
More informationThe Strong Duality Theorem 1
1/39 The Strong Duality Theorem 1 Adrian Vetta 1 This presentation is based upon the book Linear Programming by Vasek Chvatal 2/39 Part I Weak Duality 3/39 Primal and Dual Recall we have a primal linear
More informationA non-standart approach to Duality
Department of Industrial and Systems Engineering Lehigh University COR@L Seminar Series, 2005 10/27/2005 Outline 1 Duality Integration & Counting Duality for Integration Brion and Vergne s continuous formula
More informationLecture 9: Dantzig-Wolfe Decomposition
Lecture 9: Dantzig-Wolfe Decomposition (3 units) Outline Dantzig-Wolfe decomposition Column generation algorithm Relation to Lagrangian dual Branch-and-price method Generated assignment problem and multi-commodity
More informationLecture 1 Introduction
L. Vandenberghe EE236A (Fall 2013-14) Lecture 1 Introduction course overview linear optimization examples history approximate syllabus basic definitions linear optimization in vector and matrix notation
More informationDuality in Optimization and Constraint Satisfaction
Duality in Optimization and Constraint Satisfaction J. N. Hooker Carnegie Mellon University, Pittsburgh, USA john@hooker.tepper.cmu.edu Abstract. We show that various duals that occur in optimization and
More informationPart IB Optimisation
Part IB Optimisation Theorems Based on lectures by F. A. Fischer Notes taken by Dexter Chua Easter 2015 These notes are not endorsed by the lecturers, and I have modified them (often significantly) after
More informationLinear Programming. Larry Blume Cornell University, IHS Vienna and SFI. Summer 2016
Linear Programming Larry Blume Cornell University, IHS Vienna and SFI Summer 2016 These notes derive basic results in finite-dimensional linear programming using tools of convex analysis. Most sources
More informationAnswers to problems. Chapter 1. Chapter (0, 0) (3.5,0) (0,4.5) (2, 3) 2.1(a) Last tableau. (b) Last tableau /2 -3/ /4 3/4 1/4 2.
Answers to problems Chapter 1 1.1. (0, 0) (3.5,0) (0,4.5) (, 3) Chapter.1(a) Last tableau X4 X3 B /5 7/5 x -3/5 /5 Xl 4/5-1/5 8 3 Xl =,X =3,B=8 (b) Last tableau c Xl -19/ X3-3/ -7 3/4 1/4 4.5 5/4-1/4.5
More informationChap6 Duality Theory and Sensitivity Analysis
Chap6 Duality Theory and Sensitivity Analysis The rationale of duality theory Max 4x 1 + x 2 + 5x 3 + 3x 4 S.T. x 1 x 2 x 3 + 3x 4 1 5x 1 + x 2 + 3x 3 + 8x 4 55 x 1 + 2x 2 + 3x 3 5x 4 3 x 1 ~x 4 0 If we
More informationTHE EXISTENCE AND USEFULNESS OF EQUALITY CUTS IN THE MULTI-DEMAND MULTIDIMENSIONAL KNAPSACK PROBLEM LEVI DELISSA. B.S., Kansas State University, 2014
THE EXISTENCE AND USEFULNESS OF EQUALITY CUTS IN THE MULTI-DEMAND MULTIDIMENSIONAL KNAPSACK PROBLEM by LEVI DELISSA B.S., Kansas State University, 2014 A THESIS submitted in partial fulfillment of the
More informationThe Simplex Algorithm
8.433 Combinatorial Optimization The Simplex Algorithm October 6, 8 Lecturer: Santosh Vempala We proved the following: Lemma (Farkas). Let A R m n, b R m. Exactly one of the following conditions is true:.
More information1. Algebraic and geometric treatments Consider an LP problem in the standard form. x 0. Solutions to the system of linear equations
The Simplex Method Most textbooks in mathematical optimization, especially linear programming, deal with the simplex method. In this note we study the simplex method. It requires basically elementary linear
More informationA Parametric Simplex Algorithm for Linear Vector Optimization Problems
A Parametric Simplex Algorithm for Linear Vector Optimization Problems Birgit Rudloff Firdevs Ulus Robert Vanderbei July 9, 2015 Abstract In this paper, a parametric simplex algorithm for solving linear
More informationIntroduction to Integer Programming
Lecture 3/3/2006 p. /27 Introduction to Integer Programming Leo Liberti LIX, École Polytechnique liberti@lix.polytechnique.fr Lecture 3/3/2006 p. 2/27 Contents IP formulations and examples Total unimodularity
More informationMarch 2002, December Introduction. We investigate the facial structure of the convex hull of the mixed integer knapsack set
ON THE FACETS OF THE MIXED INTEGER KNAPSACK POLYHEDRON ALPER ATAMTÜRK Abstract. We study the mixed integer knapsack polyhedron, that is, the convex hull of the mixed integer set defined by an arbitrary
More informationSimplex method(s) for solving LPs in standard form
Simplex method: outline I The Simplex Method is a family of algorithms for solving LPs in standard form (and their duals) I Goal: identify an optimal basis, as in Definition 3.3 I Versions we will consider:
More informationLinear Programming Duality P&S Chapter 3 Last Revised Nov 1, 2004
Linear Programming Duality P&S Chapter 3 Last Revised Nov 1, 2004 1 In this section we lean about duality, which is another way to approach linear programming. In particular, we will see: How to define
More information