International Doctoral School Algorithmic Decision Theory: MCDA and MOO

Size: px
Start display at page:

Download "International Doctoral School Algorithmic Decision Theory: MCDA and MOO"

Transcription

1 International Doctoral School Algorithmic Decision Theory: MCDA and MOO Lecture 2: Multiobjective Linear Programming Department of Engineering Science, The University of Auckland, New Zealand Laboratoire d Informatique de Nantes Atlantique, CNRS, Université de Nantes, France MCDA and MOO, Han sur Lesse, September

2 Overview 1 Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 2 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 3 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples

3 Overview Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 1 Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 2 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 3 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples

4 Formulation and Example Solving MOLPs by Weighted Sums Variables x R n Objective function Cx where C R p n Constraints Ax = b where A R m n and b R m min {Cx : Ax = b, x 0} (1) X = {x R n : Ax = b, x 0} is the feasible set in decision space Y = {Cx : x X } is the feasible set in objective space

5 Formulation and Example Solving MOLPs by Weighted Sums Variables x R n Objective function Cx where C R p n Constraints Ax = b where A R m n and b R m min {Cx : Ax = b, x 0} (1) X = {x R n : Ax = b, x 0} is the feasible set in decision space Y = {Cx : x X } is the feasible set in objective space

6 Formulation and Example Solving MOLPs by Weighted Sums Variables x R n Objective function Cx where C R p n Constraints Ax = b where A R m n and b R m min {Cx : Ax = b, x 0} (1) X = {x R n : Ax = b, x 0} is the feasible set in decision space Y = {Cx : x X } is the feasible set in objective space

7 Formulation and Example Solving MOLPs by Weighted Sums Variables x R n Objective function Cx where C R p n Constraints Ax = b where A R m n and b R m min {Cx : Ax = b, x 0} (1) X = {x R n : Ax = b, x 0} is the feasible set in decision space Y = {Cx : x X } is the feasible set in objective space

8 Formulation and Example Solving MOLPs by Weighted Sums Example ( ) 3x1 + x min 2 x 1 2x 2 subject to x 2 3 3x 1 x 2 6 x 0 ( 3 1 C = 1 2 ) A = ( ) b = ( 3 6 )

9 Formulation and Example Solving MOLPs by Weighted Sums Example ( ) 3x1 + x min 2 x 1 2x 2 subject to x 2 3 3x 1 x 2 6 x 0 ( 3 1 C = 1 2 ) A = ( ) b = ( 3 6 )

10 Formulation and Example Solving MOLPs by Weighted Sums x 2 x 3 X x 1 x Feasible set in decision space

11 Formulation and Example Solving MOLPs by Weighted Sums Cx 1 Cx Y Cx Cx 3-10 Feasible set in objective space

12 Formulation and Example Solving MOLPs by Weighted Sums Definition Let ˆx X be a feasible solution of the MOLP (1) and let ŷ = C ˆx ˆx is called weakly efficient if there is no x X such that Cx < C ˆx; ŷ = C ˆx is called weakly nondominated ˆx is called efficient if there is no x X such that Cx C ˆx; ŷ = C ˆx is called nondominated ˆx is called properly efficient if it is efficient and if there exists a real number M > 0 such that for all i and x with c T i x < c T i ˆx there is an index j and M > 0 such that c T j x > c T j ˆx and c T i ˆx c T i x c T j x c T j ˆx M

13 Formulation and Example Solving MOLPs by Weighted Sums Definition Let ˆx X be a feasible solution of the MOLP (1) and let ŷ = C ˆx ˆx is called weakly efficient if there is no x X such that Cx < C ˆx; ŷ = C ˆx is called weakly nondominated ˆx is called efficient if there is no x X such that Cx C ˆx; ŷ = C ˆx is called nondominated ˆx is called properly efficient if it is efficient and if there exists a real number M > 0 such that for all i and x with c T i x < c T i ˆx there is an index j and M > 0 such that c T j x > c T j ˆx and c T i ˆx c T i x c T j x c T j ˆx M

14 Formulation and Example Solving MOLPs by Weighted Sums Definition Let ˆx X be a feasible solution of the MOLP (1) and let ŷ = C ˆx ˆx is called weakly efficient if there is no x X such that Cx < C ˆx; ŷ = C ˆx is called weakly nondominated ˆx is called efficient if there is no x X such that Cx C ˆx; ŷ = C ˆx is called nondominated ˆx is called properly efficient if it is efficient and if there exists a real number M > 0 such that for all i and x with c T i x < c T i ˆx there is an index j and M > 0 such that c T j x > c T j ˆx and c T i ˆx c T i x c T j x c T j ˆx M

15 Formulation and Example Solving MOLPs by Weighted Sums Definition Let ˆx X be a feasible solution of the MOLP (1) and let ŷ = C ˆx ˆx is called weakly efficient if there is no x X such that Cx < C ˆx; ŷ = C ˆx is called weakly nondominated ˆx is called efficient if there is no x X such that Cx C ˆx; ŷ = C ˆx is called nondominated ˆx is called properly efficient if it is efficient and if there exists a real number M > 0 such that for all i and x with c T i x < c T i ˆx there is an index j and M > 0 such that c T j x > c T j ˆx and c T i ˆx c T i x c T j x c T j ˆx M

16 Formulation and Example Solving MOLPs by Weighted Sums Cx 1 Cx Y Cx Cx 3-10 Nondominated set

17 Formulation and Example Solving MOLPs by Weighted Sums x 2 x 3 X x 1 x Efficient set

18 Overview Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 1 Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 2 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 3 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples

19 Formulation and Example Solving MOLPs by Weighted Sums Let λ 1,, λ p 0 and consider LP(λ) p min λ k ck T x = min λt Cx k=1 subject to Ax = b x 0 with some vector λ 0 (Why not λ = 0 or λ 0?) LP(λ) is a linear programme that can be solved by the Simplex method If λ > 0 then optimal solution of LP(λ) is properly efficient If λ 0 then optimal solution of LP(λ) is weakly efficient Converse also true, because Y convex

20 Formulation and Example Solving MOLPs by Weighted Sums Let λ 1,, λ p 0 and consider LP(λ) p min λ k ck T x = min λt Cx k=1 subject to Ax = b x 0 with some vector λ 0 (Why not λ = 0 or λ 0?) LP(λ) is a linear programme that can be solved by the Simplex method If λ > 0 then optimal solution of LP(λ) is properly efficient If λ 0 then optimal solution of LP(λ) is weakly efficient Converse also true, because Y convex

21 Formulation and Example Solving MOLPs by Weighted Sums Let λ 1,, λ p 0 and consider LP(λ) p min λ k ck T x = min λt Cx k=1 subject to Ax = b x 0 with some vector λ 0 (Why not λ = 0 or λ 0?) LP(λ) is a linear programme that can be solved by the Simplex method If λ > 0 then optimal solution of LP(λ) is properly efficient If λ 0 then optimal solution of LP(λ) is weakly efficient Converse also true, because Y convex

22 Formulation and Example Solving MOLPs by Weighted Sums Let λ 1,, λ p 0 and consider LP(λ) p min λ k ck T x = min λt Cx k=1 subject to Ax = b x 0 with some vector λ 0 (Why not λ = 0 or λ 0?) LP(λ) is a linear programme that can be solved by the Simplex method If λ > 0 then optimal solution of LP(λ) is properly efficient If λ 0 then optimal solution of LP(λ) is weakly efficient Converse also true, because Y convex

23 Formulation and Example Solving MOLPs by Weighted Sums Let λ 1,, λ p 0 and consider LP(λ) p min λ k ck T x = min λt Cx k=1 subject to Ax = b x 0 with some vector λ 0 (Why not λ = 0 or λ 0?) LP(λ) is a linear programme that can be solved by the Simplex method If λ > 0 then optimal solution of LP(λ) is properly efficient If λ 0 then optimal solution of LP(λ) is weakly efficient Converse also true, because Y convex

24 Formulation and Example Solving MOLPs by Weighted Sums Illustration in objective space Y = CX Cx 1 Cx 2 Cx 3 Cx 4 y 1 y 2 {(λ 1 ) T Cx = α 1 } {(λ 2 ) T Cx = α 2 } {(λ 3 ) T Cx = α 3 } λ 1 = (2, 1), λ 2 = (1, 3), λ 3 = (1, 1)

25 Formulation and Example Solving MOLPs by Weighted Sums Illustration in objective space Y = CX Cx 1 Cx 2 Cx 3 Cx 4 y 1 y 2 {(λ 1 ) T Cx = α 1 } {(λ 2 ) T Cx = α 2 } {(λ 3 ) T Cx = α 3 } λ 1 = (2, 1), λ 2 = (1, 3), λ 3 = (1, 1)

26 Formulation and Example Solving MOLPs by Weighted Sums y R p satisfying λ T y = α define a straight line (hyperplane) Since y = Cx and λ T Cx is minimised, we push the line towards the origin (left and down) When the line only touches Y nondominated points are found Nondominated points Y N are on the boundary of Y Y is convex polyhedron and has finite number of facets Y N consists of finitely many facets of Y The normal of the facet can serve as weight vector λ

27 Formulation and Example Solving MOLPs by Weighted Sums y R p satisfying λ T y = α define a straight line (hyperplane) Since y = Cx and λ T Cx is minimised, we push the line towards the origin (left and down) When the line only touches Y nondominated points are found Nondominated points Y N are on the boundary of Y Y is convex polyhedron and has finite number of facets Y N consists of finitely many facets of Y The normal of the facet can serve as weight vector λ

28 Formulation and Example Solving MOLPs by Weighted Sums y R p satisfying λ T y = α define a straight line (hyperplane) Since y = Cx and λ T Cx is minimised, we push the line towards the origin (left and down) When the line only touches Y nondominated points are found Nondominated points Y N are on the boundary of Y Y is convex polyhedron and has finite number of facets Y N consists of finitely many facets of Y The normal of the facet can serve as weight vector λ

29 Formulation and Example Solving MOLPs by Weighted Sums y R p satisfying λ T y = α define a straight line (hyperplane) Since y = Cx and λ T Cx is minimised, we push the line towards the origin (left and down) When the line only touches Y nondominated points are found Nondominated points Y N are on the boundary of Y Y is convex polyhedron and has finite number of facets Y N consists of finitely many facets of Y The normal of the facet can serve as weight vector λ

30 Formulation and Example Solving MOLPs by Weighted Sums y R p satisfying λ T y = α define a straight line (hyperplane) Since y = Cx and λ T Cx is minimised, we push the line towards the origin (left and down) When the line only touches Y nondominated points are found Nondominated points Y N are on the boundary of Y Y is convex polyhedron and has finite number of facets Y N consists of finitely many facets of Y The normal of the facet can serve as weight vector λ

31 Formulation and Example Solving MOLPs by Weighted Sums Question: Can all efficient solutions be found using weighted sums? If ˆx X is efficient, does there exist λ > 0 such that ˆx is optimal solution to min{λ T Cx : Ax = b, x 0}? Lemma A feasible solution x 0 X is efficient if and only if the linear programme max e T z subject to Ax = b Cx + Iz = Cx 0 x, z 0, (2) where e T = (1,, 1) R p and I is the p p identity matrix, has an optimal solution (ˆx, ẑ) with ẑ = 0

32 Formulation and Example Solving MOLPs by Weighted Sums Question: Can all efficient solutions be found using weighted sums? If ˆx X is efficient, does there exist λ > 0 such that ˆx is optimal solution to min{λ T Cx : Ax = b, x 0}? Lemma A feasible solution x 0 X is efficient if and only if the linear programme max e T z subject to Ax = b Cx + Iz = Cx 0 x, z 0, (2) where e T = (1,, 1) R p and I is the p p identity matrix, has an optimal solution (ˆx, ẑ) with ẑ = 0

33 Formulation and Example Solving MOLPs by Weighted Sums Question: Can all efficient solutions be found using weighted sums? If ˆx X is efficient, does there exist λ > 0 such that ˆx is optimal solution to min{λ T Cx : Ax = b, x 0}? Lemma A feasible solution x 0 X is efficient if and only if the linear programme max e T z subject to Ax = b Cx + Iz = Cx 0 x, z 0, (2) where e T = (1,, 1) R p and I is the p p identity matrix, has an optimal solution (ˆx, ẑ) with ẑ = 0

34 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

35 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

36 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

37 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

38 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

39 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

40 Formulation and Example Solving MOLPs by Weighted Sums Proof LP is always feasible with x = x 0, z = 0 (and value 0) Let (ˆx, ẑ) be optimal solution If ẑ = 0 then ẑ = Cx 0 C ˆx = 0 Cx 0 = C ˆx There is no x X such that Cx Cx 0 because (x, Cx 0 Cx) would be better solution x 0 efficient If ˆx 0 efficient there is no x X with Cx Cx 0 there is no z with z = Cx 0 Cx 0 max e T z 0 max e T z = 0

41 Formulation and Example Solving MOLPs by Weighted Sums Lemma A feasible solution x 0 X is efficient if and only if the linear programme min u T b + w T Cx 0 subject to u T A + w T C 0 w e u R m (3) has an optimal solution (û, ŵ) with û T b + ŵ T Cx 0 = 0 Proof The LP (3) is the dual of the LP (2)

42 Formulation and Example Solving MOLPs by Weighted Sums Lemma A feasible solution x 0 X is efficient if and only if the linear programme min u T b + w T Cx 0 subject to u T A + w T C 0 w e u R m (3) has an optimal solution (û, ŵ) with û T b + ŵ T Cx 0 = 0 Proof The LP (3) is the dual of the LP (2)

43 Formulation and Example Solving MOLPs by Weighted Sums Theorem A feasible solution x 0 X is an efficient solution of the MOLP (1) if and only if there exists a λ R p > such that for all x X λ T Cx 0 λ T Cx (4) Note: We already know that optimal solutions of weighted sum problems are efficient

44 Formulation and Example Solving MOLPs by Weighted Sums Theorem A feasible solution x 0 X is an efficient solution of the MOLP (1) if and only if there exists a λ R p > such that for all x X λ T Cx 0 λ T Cx (4) Note: We already know that optimal solutions of weighted sum problems are efficient

45 Formulation and Example Solving MOLPs by Weighted Sums Proof Let x 0 X E By Lemma 4 LP (3) has an optimal solution (û, ŵ) such that û T b = ŵ T Cx 0 (5) û is also an optimal solution of the LP { } min u T b : u T A ŵ T C, (6) which is (3) with w = ŵ fixed There is an optimal solution of the dual of (6) { } max ŵ T Cx : Ax = b, x 0 (7)

46 Formulation and Example Solving MOLPs by Weighted Sums Proof Let x 0 X E By Lemma 4 LP (3) has an optimal solution (û, ŵ) such that û T b = ŵ T Cx 0 (5) û is also an optimal solution of the LP { } min u T b : u T A ŵ T C, (6) which is (3) with w = ŵ fixed There is an optimal solution of the dual of (6) { } max ŵ T Cx : Ax = b, x 0 (7)

47 Formulation and Example Solving MOLPs by Weighted Sums Proof Let x 0 X E By Lemma 4 LP (3) has an optimal solution (û, ŵ) such that û T b = ŵ T Cx 0 (5) û is also an optimal solution of the LP { } min u T b : u T A ŵ T C, (6) which is (3) with w = ŵ fixed There is an optimal solution of the dual of (6) { } max ŵ T Cx : Ax = b, x 0 (7)

48 Formulation and Example Solving MOLPs by Weighted Sums Proof Let x 0 X E By Lemma 4 LP (3) has an optimal solution (û, ŵ) such that û T b = ŵ T Cx 0 (5) û is also an optimal solution of the LP { } min u T b : u T A ŵ T C, (6) which is (3) with w = ŵ fixed There is an optimal solution of the dual of (6) { } max ŵ T Cx : Ax = b, x 0 (7)

49 Formulation and Example Solving MOLPs by Weighted Sums Proof By weak duality u T b ŵ T Cx for all feasible solutions u of (6) and for all feasible solutions x of (7) We already know that û T b = ŵ T Cx 0 from (5) x 0 is an optimal solution of (7) Note that (7) is equivalent to { } min ŵ T Cx : Ax = b, x 0 with ŵ e > 0 from the constraints in (3)

50 Formulation and Example Solving MOLPs by Weighted Sums Proof By weak duality u T b ŵ T Cx for all feasible solutions u of (6) and for all feasible solutions x of (7) We already know that û T b = ŵ T Cx 0 from (5) x 0 is an optimal solution of (7) Note that (7) is equivalent to { } min ŵ T Cx : Ax = b, x 0 with ŵ e > 0 from the constraints in (3)

51 Formulation and Example Solving MOLPs by Weighted Sums Proof By weak duality u T b ŵ T Cx for all feasible solutions u of (6) and for all feasible solutions x of (7) We already know that û T b = ŵ T Cx 0 from (5) x 0 is an optimal solution of (7) Note that (7) is equivalent to { } min ŵ T Cx : Ax = b, x 0 with ŵ e > 0 from the constraints in (3)

52 Formulation and Example Solving MOLPs by Weighted Sums Proof By weak duality u T b ŵ T Cx for all feasible solutions u of (6) and for all feasible solutions x of (7) We already know that û T b = ŵ T Cx 0 from (5) x 0 is an optimal solution of (7) Note that (7) is equivalent to { } min ŵ T Cx : Ax = b, x 0 with ŵ e > 0 from the constraints in (3)

53 Overview Multiobjective Linear Programming The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 1 Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 2 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 3 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples

54 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Modification of the Simplex algorithm for LPs with two objectives min ( (c 1 ) T x, (c 2 ) T x ) subject to Ax = b x 0 (8) We can find all efficient solutions by solving the parametric LP { } min λ 1 (c 1 ) T x + λ 2 (c 2 ) T x : Ax = b, x 0 for all λ = (λ 1, λ 2 ) > 0

55 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Modification of the Simplex algorithm for LPs with two objectives min ( (c 1 ) T x, (c 2 ) T x ) subject to Ax = b x 0 (8) We can find all efficient solutions by solving the parametric LP { } min λ 1 (c 1 ) T x + λ 2 (c 2 ) T x : Ax = b, x 0 for all λ = (λ 1, λ 2 ) > 0

56 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example We can divide the objective by λ 1 + λ 2 without changing the optima, ie λ 1 = λ 1/(λ 1 + λ 2 ), λ 2 = λ 2/(λ 1 + λ 2 and λ 1 + λ 2 = 1 or λ 2 = 1 λ 1 LPs with one parameter 0 λ 1 and parametric objective c(λ) := λc 1 + (1 λ)c 2 { } min c(λ) T x : Ax = b, x 0 (9)

57 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example We can divide the objective by λ 1 + λ 2 without changing the optima, ie λ 1 = λ 1/(λ 1 + λ 2 ), λ 2 = λ 2/(λ 1 + λ 2 and λ 1 + λ 2 = 1 or λ 2 = 1 λ 1 LPs with one parameter 0 λ 1 and parametric objective c(λ) := λc 1 + (1 λ)c 2 { } min c(λ) T x : Ax = b, x 0 (9)

58 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Let B be a feasible basis Recall reduced cost c N = c N cb T B 1 N Reduced cost for the parametric LP c(λ) = λ c 1 + (1 λ) c 2 (10) Suppose ˆB is an optimal basis of (9) for some ˆλ c(ˆλ) 0

59 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Let B be a feasible basis Recall reduced cost c N = c N cb T B 1 N Reduced cost for the parametric LP c(λ) = λ c 1 + (1 λ) c 2 (10) Suppose ˆB is an optimal basis of (9) for some ˆλ c(ˆλ) 0

60 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Let B be a feasible basis Recall reduced cost c N = c N cb T B 1 N Reduced cost for the parametric LP c(λ) = λ c 1 + (1 λ) c 2 (10) Suppose ˆB is an optimal basis of (9) for some ˆλ c(ˆλ) 0

61 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Let B be a feasible basis Recall reduced cost c N = c N cb T B 1 N Reduced cost for the parametric LP c(λ) = λ c 1 + (1 λ) c 2 (10) Suppose ˆB is an optimal basis of (9) for some ˆλ c(ˆλ) 0

62 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Let B be a feasible basis Recall reduced cost c N = c N cb T B 1 N Reduced cost for the parametric LP c(λ) = λ c 1 + (1 λ) c 2 (10) Suppose ˆB is an optimal basis of (9) for some ˆλ c(ˆλ) 0

63 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

64 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

65 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

66 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

67 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

68 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

69 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

70 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

71 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Case 1: c 2 0 From (10) c(λ) 0 for all λ < ˆλ ˆB is optimal basis for all 0 λ ˆλ Case 2: There is at least one i N with c 2 i < 0 there is λ < ˆλ such that c(λ) i = 0 λ c 1 i + (1 λ) c 2 i = 0 λ( c 1 i c 2 i ) + c2 i = 0 c2 i λ = c i 1 c2 i Below this value ˆB is not optimal

72 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example I = {i N : c 2 i < 0, c 1 i 0} λ := max i I c 1 i c i 2 c i 2 (11) ˆB is optimal for all λ [λ, ˆλ] As soon as λ < λ new bases become optimal Entering variable x s has to be chosen where the maximum in (11) is attained for i = s

73 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example I = {i N : c 2 i < 0, c 1 i 0} λ := max i I c 1 i c i 2 c i 2 (11) ˆB is optimal for all λ [λ, ˆλ] As soon as λ < λ new bases become optimal Entering variable x s has to be chosen where the maximum in (11) is attained for i = s

74 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example I = {i N : c 2 i < 0, c 1 i 0} λ := max i I c 1 i c i 2 c i 2 (11) ˆB is optimal for all λ [λ, ˆλ] As soon as λ < λ new bases become optimal Entering variable x s has to be chosen where the maximum in (11) is attained for i = s

75 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example I = {i N : c 2 i < 0, c 1 i 0} λ := max i I c 1 i c i 2 c i 2 (11) ˆB is optimal for all λ [λ, ˆλ] As soon as λ < λ new bases become optimal Entering variable x s has to be chosen where the maximum in (11) is attained for i = s

76 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example I = {i N : c 2 i < 0, c 1 i 0} λ := max i I c 1 i c i 2 c i 2 (11) ˆB is optimal for all λ [λ, ˆλ] As soon as λ < λ new bases become optimal Entering variable x s has to be chosen where the maximum in (11) is attained for i = s

77 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Algorithm (Parametric Simplex for biobjective LPs) Input: Data A, b, C for a biobjective LP Phase I: Solve the auxiliary LP for Phase I using the Simplex algorithm If the optimal value is positive, STOP, X = Otherwise let B be an optimal basis Phase II: Solve the LP (9) for λ = 1 starting from basis B found in Phase I yielding an optimal basis ˆB Compute à and b Phase III: While I = {i N : c 2 i < 0, c 1 i 0} = c 2 i λ := max i I c i 1 c2 i n o c i s argmax i I : 2 c n i 1 c2 i o b r argmin j B : j, à à js > 0 js Let B := (B \ {r}) {s} and update à and b End while Output: Sequence of λ-values and sequence of optimal BFSs

78 Overview Multiobjective Linear Programming The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 1 Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 2 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 3 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples

79 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Example LP(λ) min ( 3x1 + x 2 ) x 1 2x 2 subject to x 2 3 3x 1 x 2 6 x 0 min (4λ 1)x 1 + (3λ 2)x 2 subject to x 2 + x 3 = 3 3x 1 x 2 + x 4 = 6 x 0

80 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Use Simplex tableaus showing reduced cost vectors c 1 and c 2 Optimal basis for λ = 1 is B = {3, 4}, optimal basic feasible solution x = (0, 0, 3, 6) Start with Phase 3

81 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Use Simplex tableaus showing reduced cost vectors c 1 and c 2 Optimal basis for λ = 1 is B = {3, 4}, optimal basic feasible solution x = (0, 0, 3, 6) Start with Phase 3

82 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Use Simplex tableaus showing reduced cost vectors c 1 and c 2 Optimal basis for λ = 1 is B = {3, 4}, optimal basic feasible solution x = (0, 0, 3, 6) Start with Phase 3

83 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Iteration 1: c c x x λ = 1, c(λ) = (3, 1, 0, { 0), B 1 = {3, } 4}, x 1 = (0, 0, 3, 6) I = {1, 2}, λ = max 1 3+1, = 2 3 s = 2, r = 3

84 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Iteration 2 c c x x λ = 2/3, c(λ) = (5/3, { 0, 0, } 0), B 2 = {2, 4}, x 2 = (0, 3, 0, 9) I = {1}, λ = max = 1 4 s = 1, r = 4

85 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Iteration 3 c c /3 1/3 9 x x /3 1/3 3 λ = 1/4, c(λ) = (0, 0, 5/4, 0), B 3 = {1, 2}, x 3 = (3, 3, 0, 0) I =

86 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

87 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

88 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

89 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

90 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

91 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

92 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight values λ 1 = 1, λ 2 = 2/3, λ 3 = 1/4, λ 4 = 0 Basic feasible solutions x 1, x 2, x 3 In each iteration c(λ) can be calculated with the previous and current c 1 and c 2 Basis B 1 = (3, 4) and BFS x 1 = (0, 0, 3, 6) are optimal for λ [2/3, 1] Basis B 2 = (2, 4) and BFS x 2 = (0, 3, 0, 9) are optimal for λ [1/4, 2/3], and Basis B 3 = (1, 2) and BFS x 3 = (3, 3, 0, 0) are optimal for λ [0, 1/4] Objective vectors for basic feasible solutions: Cx 1 = (0, 0), Cx 2 = (3, 6), and Cx 3 = (12, 9)

93 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Values λ = 2/3 and λ = 1/4 correspond to weight vectors (2/3, 1/3) and (1/4, 3/4) Contour lines for weighted sum objectives in decision are parallel to efficient edges 2 3 (3x 1 + x 2 ) ( x 1 2x 2 ) = 5 3 x (3x 1 + x 2 ) ( x 1 2x 2 ) = 5 4 x x 2 x 3 X Feasible set in decision space and efficient set 1 x 1 x

94 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Values λ = 2/3 and λ = 1/4 correspond to weight vectors (2/3, 1/3) and (1/4, 3/4) Contour lines for weighted sum objectives in decision are parallel to efficient edges 2 3 (3x 1 + x 2 ) ( x 1 2x 2 ) = 5 3 x (3x 1 + x 2 ) ( x 1 2x 2 ) = 5 4 x x 2 x 3 X Feasible set in decision space and efficient set 1 x 1 x

95 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Values λ = 2/3 and λ = 1/4 correspond to weight vectors (2/3, 1/3) and (1/4, 3/4) Contour lines for weighted sum objectives in decision are parallel to efficient edges 2 3 (3x 1 + x 2 ) ( x 1 2x 2 ) = 5 3 x (3x 1 + x 2 ) ( x 1 2x 2 ) = 5 4 x x 2 x 3 X Feasible set in decision space and efficient set 1 x 1 x

96 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Weight vectors (2/3, 1/3) and (1/4, 3/4) are normal to nondominated edges Cx 1 Cx Y Cx Cx 3-10 Objective space and nondominated set

97 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Example Algorithm finds all nondominated extreme points in objective space and one efficient bfs for each of those Algorithm does not find all efficient solutions just as Simplex algorithm does not find all optimal solutions of an LP min (x 1, x 2 ) T subject to 0 x i 1 i = 1, 2, 3 Efficient set: {x R 3 : x 1 = x 2 = 0, 0 x 3 1}

98 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Example Algorithm finds all nondominated extreme points in objective space and one efficient bfs for each of those Algorithm does not find all efficient solutions just as Simplex algorithm does not find all optimal solutions of an LP min (x 1, x 2 ) T subject to 0 x i 1 i = 1, 2, 3 Efficient set: {x R 3 : x 1 = x 2 = 0, 0 x 3 1}

99 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example Example Algorithm finds all nondominated extreme points in objective space and one efficient bfs for each of those Algorithm does not find all efficient solutions just as Simplex algorithm does not find all optimal solutions of an LP min (x 1, x 2 ) T subject to 0 x i 1 i = 1, 2, 3 Efficient set: {x R 3 : x 1 = x 2 = 0, 0 x 3 1}

100 Overview Multiobjective Linear Programming A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples 1 Multiobjective Linear Programming Formulation and Example Solving MOLPs by Weighted Sums 2 The Parametric Simplex Algorithm Biobjective Linear Programmes: Example 3 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples

101 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples min{cx : Ax = b, x 0} Let B be a basis and C = C C B A 1 B A and R = C N How to calculate critical λ if p > 2? ( ) 3 1 At B 1 : C N =, λ 1 2 = 2/3, λ = (2/3, 1/3) T and λ T C N = (5/3, 0) T ( At B 2 : C 3 1 N = 1 2 λ T CN = (0, 5/4) T ), λ = 1/4, λ = (1/4, 3/4) T and Find λ R p, λ > 0 such that λ T R 0 (optimality) and λ T r j = 0 (alternative optimum) for some column r j of R

102 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples min{cx : Ax = b, x 0} Let B be a basis and C = C C B A 1 B A and R = C N How to calculate critical λ if p > 2? ( ) 3 1 At B 1 : C N =, λ 1 2 = 2/3, λ = (2/3, 1/3) T and λ T C N = (5/3, 0) T ( At B 2 : C 3 1 N = 1 2 λ T CN = (0, 5/4) T ), λ = 1/4, λ = (1/4, 3/4) T and Find λ R p, λ > 0 such that λ T R 0 (optimality) and λ T r j = 0 (alternative optimum) for some column r j of R

103 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples min{cx : Ax = b, x 0} Let B be a basis and C = C C B A 1 B A and R = C N How to calculate critical λ if p > 2? ( ) 3 1 At B 1 : C N =, λ 1 2 = 2/3, λ = (2/3, 1/3) T and λ T C N = (5/3, 0) T ( At B 2 : C 3 1 N = 1 2 λ T CN = (0, 5/4) T ), λ = 1/4, λ = (1/4, 3/4) T and Find λ R p, λ > 0 such that λ T R 0 (optimality) and λ T r j = 0 (alternative optimum) for some column r j of R

104 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples min{cx : Ax = b, x 0} Let B be a basis and C = C C B A 1 B A and R = C N How to calculate critical λ if p > 2? ( ) 3 1 At B 1 : C N =, λ 1 2 = 2/3, λ = (2/3, 1/3) T and λ T C N = (5/3, 0) T ( At B 2 : C 3 1 N = 1 2 λ T CN = (0, 5/4) T ), λ = 1/4, λ = (1/4, 3/4) T and Find λ R p, λ > 0 such that λ T R 0 (optimality) and λ T r j = 0 (alternative optimum) for some column r j of R

105 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples min{cx : Ax = b, x 0} Let B be a basis and C = C C B A 1 B A and R = C N How to calculate critical λ if p > 2? ( ) 3 1 At B 1 : C N =, λ 1 2 = 2/3, λ = (2/3, 1/3) T and λ T C N = (5/3, 0) T ( At B 2 : C 3 1 N = 1 2 λ T CN = (0, 5/4) T ), λ = 1/4, λ = (1/4, 3/4) T and Find λ R p, λ > 0 such that λ T R 0 (optimality) and λ T r j = 0 (alternative optimum) for some column r j of R

106 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples min{cx : Ax = b, x 0} Let B be a basis and C = C C B A 1 B A and R = C N How to calculate critical λ if p > 2? ( ) 3 1 At B 1 : C N =, λ 1 2 = 2/3, λ = (2/3, 1/3) T and λ T C N = (5/3, 0) T ( At B 2 : C 3 1 N = 1 2 λ T CN = (0, 5/4) T ), λ = 1/4, λ = (1/4, 3/4) T and Find λ R p, λ > 0 such that λ T R 0 (optimality) and λ T r j = 0 (alternative optimum) for some column r j of R

107 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma If X E then X has an efficient basic feasible solution Proof There is some λ > 0 such that min x X λ T Cx has an optimal solution Thus LP(λ) has an optimal basic feasible solution solution, which is an efficient solution of the MOLP

108 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma If X E then X has an efficient basic feasible solution Proof There is some λ > 0 such that min x X λ T Cx has an optimal solution Thus LP(λ) has an optimal basic feasible solution solution, which is an efficient solution of the MOLP

109 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma If X E then X has an efficient basic feasible solution Proof There is some λ > 0 such that min x X λ T Cx has an optimal solution Thus LP(λ) has an optimal basic feasible solution solution, which is an efficient solution of the MOLP

110 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Definition 1 A feasible basis B is called efficient basis if B is an optimal basis of LP(λ) for some λ R p > 2 Two bases B and ˆB are called adjacent if one can be obtained from the other by a single pivot step 3 Let B be an efficient basis Variable x j, j N is called efficient nonbasic variable at B if there exists a λ R p > such that λ T R 0 and λ T r j = 0, where r j is the column of R corresponding to variable x j 4 Let B be an efficient basis and let x j be an efficient nonbasic variable Then a feasible pivot from B with x j entering the basis (even with negative pivot element) is called an efficient pivot with respect to B and x j

111 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Definition 1 A feasible basis B is called efficient basis if B is an optimal basis of LP(λ) for some λ R p > 2 Two bases B and ˆB are called adjacent if one can be obtained from the other by a single pivot step 3 Let B be an efficient basis Variable x j, j N is called efficient nonbasic variable at B if there exists a λ R p > such that λ T R 0 and λ T r j = 0, where r j is the column of R corresponding to variable x j 4 Let B be an efficient basis and let x j be an efficient nonbasic variable Then a feasible pivot from B with x j entering the basis (even with negative pivot element) is called an efficient pivot with respect to B and x j

112 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Definition 1 A feasible basis B is called efficient basis if B is an optimal basis of LP(λ) for some λ R p > 2 Two bases B and ˆB are called adjacent if one can be obtained from the other by a single pivot step 3 Let B be an efficient basis Variable x j, j N is called efficient nonbasic variable at B if there exists a λ R p > such that λ T R 0 and λ T r j = 0, where r j is the column of R corresponding to variable x j 4 Let B be an efficient basis and let x j be an efficient nonbasic variable Then a feasible pivot from B with x j entering the basis (even with negative pivot element) is called an efficient pivot with respect to B and x j

113 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Definition 1 A feasible basis B is called efficient basis if B is an optimal basis of LP(λ) for some λ R p > 2 Two bases B and ˆB are called adjacent if one can be obtained from the other by a single pivot step 3 Let B be an efficient basis Variable x j, j N is called efficient nonbasic variable at B if there exists a λ R p > such that λ T R 0 and λ T r j = 0, where r j is the column of R corresponding to variable x j 4 Let B be an efficient basis and let x j be an efficient nonbasic variable Then a feasible pivot from B with x j entering the basis (even with negative pivot element) is called an efficient pivot with respect to B and x j

114 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples No efficient basis is optimal for all p objectives at the same time Therefore R always contains positive and negative entries Proposition Let B be an efficient basis There exists an efficient nonbasic variable at B

115 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples No efficient basis is optimal for all p objectives at the same time Therefore R always contains positive and negative entries Proposition Let B be an efficient basis There exists an efficient nonbasic variable at B

116 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples No efficient basis is optimal for all p objectives at the same time Therefore R always contains positive and negative entries Proposition Let B be an efficient basis There exists an efficient nonbasic variable at B

117 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples It is not possible to define efficient nonbasic variables by the existence of a column in R with positive and negative entries Example R = ( ) λ T r 2 = 0 requires λ 2 = 2λ 1 λ T r 1 0 requires λ 1 0, an impossibility for λ > 0

118 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

119 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

120 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

121 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

122 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

123 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

124 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Lemma Let B be an efficient basis and x j be an efficient nonbasic variable Then any efficient pivot from B leads to an adjacent efficient basis ˆB Proof x j efficient entering variable at basis B there is λ R p > with λ T R 0 and λ T r j = 0 x j is nonbasic variable with reduced cost 0 in LP(λ) Reduced costs of LP(λ) do not change after a pivot with x j entering Let ˆB be the resulting basis with feasible pivot and x j entering Because λ T R 0 and λ T r j = 0 at ˆB, ˆB is an optimal basis for LP(λ) and therefore an adjacent efficient basis

125 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples How to identify efficient nonbasic variables? Theorem Let B be an efficient basis and let x j be a nonbasic variable Variable x j is an efficient nonbasic variable if and only if the LP has an optimal value of 0 max e t v subject to Rz r j δ + Iv = 0 z, δ, v 0 (12) (12) is always feasible with (z, δ, v) = 0

126 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples How to identify efficient nonbasic variables? Theorem Let B be an efficient basis and let x j be a nonbasic variable Variable x j is an efficient nonbasic variable if and only if the LP has an optimal value of 0 max e t v subject to Rz r j δ + Iv = 0 z, δ, v 0 (12) (12) is always feasible with (z, δ, v) = 0

127 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Proof By definition x j is an efficient nonbasic variable if the LP min 0 T λ = 0 subject to R T λ 0 (r j ) T λ = 0 I λ e (13) has an optimal objective value of 0, ie if it is feasible (13) is equivalent to min 0 T λ = 0 subject to R T λ 0 (r j ) T λ 0 I λ e (14)

128 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Proof By definition x j is an efficient nonbasic variable if the LP min 0 T λ = 0 subject to R T λ 0 (r j ) T λ = 0 I λ e (13) has an optimal objective value of 0, ie if it is feasible (13) is equivalent to min 0 T λ = 0 subject to R T λ 0 (r j ) T λ 0 I λ e (14)

129 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Proof The dual of (14) is max e T v subject to Rz r j δ + Iv = 0 z, δ, v 0 (15)

130 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Need to show: ALL efficient bases can be reached by efficient pivots Definition Two efficient bases B and ˆB are called connected if one can be obtained from the other by performing only efficient pivots Theorem All efficient bases are connected

131 A Multiobjective Simplex Algorithm Multiobjective Simplex: Examples Need to show: ALL efficient bases can be reached by efficient pivots Definition Two efficient bases B and ˆB are called connected if one can be obtained from the other by performing only efficient pivots Theorem All efficient bases are connected

1 Introduction. Linear Programming. Questions. A general optimization problem is of the form: choose x to. max f(x) subject to x S. where.

1 Introduction. Linear Programming. Questions. A general optimization problem is of the form: choose x to. max f(x) subject to x S. where. Introduction Linear Programming Neil Laws TT 00 A general optimization problem is of the form: choose x to maximise f(x) subject to x S where x = (x,..., x n ) T, f : R n R is the objective function, S

More information

. P. 4.3 Basic feasible solutions and vertices of polyhedra. x 1. x 2

. P. 4.3 Basic feasible solutions and vertices of polyhedra. x 1. x 2 4. Basic feasible solutions and vertices of polyhedra Due to the fundamental theorem of Linear Programming, to solve any LP it suffices to consider the vertices (finitely many) of the polyhedron P of the

More information

1 Solving LPs: The Simplex Algorithm of George Dantzig

1 Solving LPs: The Simplex Algorithm of George Dantzig Solving LPs: The Simplex Algorithm of George Dantzig. Simplex Pivoting: Dictionary Format We illustrate a general solution procedure, called the simplex algorithm, by implementing it on a very simple example.

More information

Duality in Linear Programming

Duality in Linear Programming Duality in Linear Programming 4 In the preceding chapter on sensitivity analysis, we saw that the shadow-price interpretation of the optimal simplex multipliers is a very useful concept. First, these shadow

More information

2.3 Convex Constrained Optimization Problems

2.3 Convex Constrained Optimization Problems 42 CHAPTER 2. FUNDAMENTAL CONCEPTS IN CONVEX OPTIMIZATION Theorem 15 Let f : R n R and h : R R. Consider g(x) = h(f(x)) for all x R n. The function g is convex if either of the following two conditions

More information

Optimization in R n Introduction

Optimization in R n Introduction Optimization in R n Introduction Rudi Pendavingh Eindhoven Technical University Optimization in R n, lecture Rudi Pendavingh (TUE) Optimization in R n Introduction ORN / 4 Some optimization problems designing

More information

4.6 Linear Programming duality

4.6 Linear Programming duality 4.6 Linear Programming duality To any minimization (maximization) LP we can associate a closely related maximization (minimization) LP. Different spaces and objective functions but in general same optimal

More information

Lecture 3. Linear Programming. 3B1B Optimization Michaelmas 2015 A. Zisserman. Extreme solutions. Simplex method. Interior point method

Lecture 3. Linear Programming. 3B1B Optimization Michaelmas 2015 A. Zisserman. Extreme solutions. Simplex method. Interior point method Lecture 3 3B1B Optimization Michaelmas 2015 A. Zisserman Linear Programming Extreme solutions Simplex method Interior point method Integer programming and relaxation The Optimization Tree Linear Programming

More information

Linear Programming in Matrix Form

Linear Programming in Matrix Form Linear Programming in Matrix Form Appendix B We first introduce matrix concepts in linear programming by developing a variation of the simplex method called the revised simplex method. This algorithm,

More information

Practical Guide to the Simplex Method of Linear Programming

Practical Guide to the Simplex Method of Linear Programming Practical Guide to the Simplex Method of Linear Programming Marcel Oliver Revised: April, 0 The basic steps of the simplex algorithm Step : Write the linear programming problem in standard form Linear

More information

LECTURE 5: DUALITY AND SENSITIVITY ANALYSIS. 1. Dual linear program 2. Duality theory 3. Sensitivity analysis 4. Dual simplex method

LECTURE 5: DUALITY AND SENSITIVITY ANALYSIS. 1. Dual linear program 2. Duality theory 3. Sensitivity analysis 4. Dual simplex method LECTURE 5: DUALITY AND SENSITIVITY ANALYSIS 1. Dual linear program 2. Duality theory 3. Sensitivity analysis 4. Dual simplex method Introduction to dual linear program Given a constraint matrix A, right

More information

Mathematical finance and linear programming (optimization)

Mathematical finance and linear programming (optimization) Mathematical finance and linear programming (optimization) Geir Dahl September 15, 2009 1 Introduction The purpose of this short note is to explain how linear programming (LP) (=linear optimization) may

More information

Linear Programming. March 14, 2014

Linear Programming. March 14, 2014 Linear Programming March 1, 01 Parts of this introduction to linear programming were adapted from Chapter 9 of Introduction to Algorithms, Second Edition, by Cormen, Leiserson, Rivest and Stein [1]. 1

More information

Chapter 6. Linear Programming: The Simplex Method. Introduction to the Big M Method. Section 4 Maximization and Minimization with Problem Constraints

Chapter 6. Linear Programming: The Simplex Method. Introduction to the Big M Method. Section 4 Maximization and Minimization with Problem Constraints Chapter 6 Linear Programming: The Simplex Method Introduction to the Big M Method In this section, we will present a generalized version of the simplex method that t will solve both maximization i and

More information

Some representability and duality results for convex mixed-integer programs.

Some representability and duality results for convex mixed-integer programs. Some representability and duality results for convex mixed-integer programs. Santanu S. Dey Joint work with Diego Morán and Juan Pablo Vielma December 17, 2012. Introduction About Motivation Mixed integer

More information

Simplex method summary

Simplex method summary Simplex method summary Problem: optimize a linear objective, subject to linear constraints 1. Step 1: Convert to standard form: variables on right-hand side, positive constant on left slack variables for

More information

1 Linear Programming. 1.1 Introduction. Problem description: motivate by min-cost flow. bit of history. everything is LP. NP and conp. P breakthrough.

1 Linear Programming. 1.1 Introduction. Problem description: motivate by min-cost flow. bit of history. everything is LP. NP and conp. P breakthrough. 1 Linear Programming 1.1 Introduction Problem description: motivate by min-cost flow bit of history everything is LP NP and conp. P breakthrough. general form: variables constraints: linear equalities

More information

Integer Programming: Algorithms - 3

Integer Programming: Algorithms - 3 Week 9 Integer Programming: Algorithms - 3 OPR 992 Applied Mathematical Programming OPR 992 - Applied Mathematical Programming - p. 1/12 Dantzig-Wolfe Reformulation Example Strength of the Linear Programming

More information

Systems of Linear Equations

Systems of Linear Equations Systems of Linear Equations Beifang Chen Systems of linear equations Linear systems A linear equation in variables x, x,, x n is an equation of the form a x + a x + + a n x n = b, where a, a,, a n and

More information

Lecture 2: August 29. Linear Programming (part I)

Lecture 2: August 29. Linear Programming (part I) 10-725: Convex Optimization Fall 2013 Lecture 2: August 29 Lecturer: Barnabás Póczos Scribes: Samrachana Adhikari, Mattia Ciollaro, Fabrizio Lecci Note: LaTeX template courtesy of UC Berkeley EECS dept.

More information

CHAPTER 9. Integer Programming

CHAPTER 9. Integer Programming CHAPTER 9 Integer Programming An integer linear program (ILP) is, by definition, a linear program with the additional constraint that all variables take integer values: (9.1) max c T x s t Ax b and x integral

More information

Linear Programming I

Linear Programming I Linear Programming I November 30, 2003 1 Introduction In the VCR/guns/nuclear bombs/napkins/star wars/professors/butter/mice problem, the benevolent dictator, Bigus Piguinus, of south Antarctica penguins

More information

Degeneracy in Linear Programming

Degeneracy in Linear Programming Degeneracy in Linear Programming I heard that today s tutorial is all about Ellen DeGeneres Sorry, Stan. But the topic is just as interesting. It s about degeneracy in Linear Programming. Degeneracy? Students

More information

IEOR 4404 Homework #2 Intro OR: Deterministic Models February 14, 2011 Prof. Jay Sethuraman Page 1 of 5. Homework #2

IEOR 4404 Homework #2 Intro OR: Deterministic Models February 14, 2011 Prof. Jay Sethuraman Page 1 of 5. Homework #2 IEOR 4404 Homework # Intro OR: Deterministic Models February 14, 011 Prof. Jay Sethuraman Page 1 of 5 Homework #.1 (a) What is the optimal solution of this problem? Let us consider that x 1, x and x 3

More information

Special Situations in the Simplex Algorithm

Special Situations in the Simplex Algorithm Special Situations in the Simplex Algorithm Degeneracy Consider the linear program: Maximize 2x 1 +x 2 Subject to: 4x 1 +3x 2 12 (1) 4x 1 +x 2 8 (2) 4x 1 +2x 2 8 (3) x 1, x 2 0. We will first apply the

More information

Linear Programming: Theory and Applications

Linear Programming: Theory and Applications Linear Programming: Theory and Applications Catherine Lewis May 11, 2008 1 Contents 1 Introduction to Linear Programming 3 1.1 What is a linear program?...................... 3 1.2 Assumptions.............................

More information

An Introduction to Linear Programming

An Introduction to Linear Programming An Introduction to Linear Programming Steven J. Miller March 31, 2007 Mathematics Department Brown University 151 Thayer Street Providence, RI 02912 Abstract We describe Linear Programming, an important

More information

Proximal mapping via network optimization

Proximal mapping via network optimization L. Vandenberghe EE236C (Spring 23-4) Proximal mapping via network optimization minimum cut and maximum flow problems parametric minimum cut problem application to proximal mapping Introduction this lecture:

More information

Lecture 5 Principal Minors and the Hessian

Lecture 5 Principal Minors and the Hessian Lecture 5 Principal Minors and the Hessian Eivind Eriksen BI Norwegian School of Management Department of Economics October 01, 2010 Eivind Eriksen (BI Dept of Economics) Lecture 5 Principal Minors and

More information

Linear Programming. Widget Factory Example. Linear Programming: Standard Form. Widget Factory Example: Continued.

Linear Programming. Widget Factory Example. Linear Programming: Standard Form. Widget Factory Example: Continued. Linear Programming Widget Factory Example Learning Goals. Introduce Linear Programming Problems. Widget Example, Graphical Solution. Basic Theory:, Vertices, Existence of Solutions. Equivalent formulations.

More information

5.1 Bipartite Matching

5.1 Bipartite Matching CS787: Advanced Algorithms Lecture 5: Applications of Network Flow In the last lecture, we looked at the problem of finding the maximum flow in a graph, and how it can be efficiently solved using the Ford-Fulkerson

More information

Reduced echelon form: Add the following conditions to conditions 1, 2, and 3 above:

Reduced echelon form: Add the following conditions to conditions 1, 2, and 3 above: Section 1.2: Row Reduction and Echelon Forms Echelon form (or row echelon form): 1. All nonzero rows are above any rows of all zeros. 2. Each leading entry (i.e. left most nonzero entry) of a row is in

More information

Chapter 4. Duality. 4.1 A Graphical Example

Chapter 4. Duality. 4.1 A Graphical Example Chapter 4 Duality Given any linear program, there is another related linear program called the dual. In this chapter, we will develop an understanding of the dual linear program. This understanding translates

More information

Nonlinear Programming Methods.S2 Quadratic Programming

Nonlinear Programming Methods.S2 Quadratic Programming Nonlinear Programming Methods.S2 Quadratic Programming Operations Research Models and Methods Paul A. Jensen and Jonathan F. Bard A linearly constrained optimization problem with a quadratic objective

More information

Applied Algorithm Design Lecture 5

Applied Algorithm Design Lecture 5 Applied Algorithm Design Lecture 5 Pietro Michiardi Eurecom Pietro Michiardi (Eurecom) Applied Algorithm Design Lecture 5 1 / 86 Approximation Algorithms Pietro Michiardi (Eurecom) Applied Algorithm Design

More information

This exposition of linear programming

This exposition of linear programming Linear Programming and the Simplex Method David Gale This exposition of linear programming and the simplex method is intended as a companion piece to the article in this issue on the life and work of George

More information

Chapter 2 Solving Linear Programs

Chapter 2 Solving Linear Programs Chapter 2 Solving Linear Programs Companion slides of Applied Mathematical Programming by Bradley, Hax, and Magnanti (Addison-Wesley, 1977) prepared by José Fernando Oliveira Maria Antónia Carravilla A

More information

56:171 Operations Research Midterm Exam Solutions Fall 2001

56:171 Operations Research Midterm Exam Solutions Fall 2001 56:171 Operations Research Midterm Exam Solutions Fall 2001 True/False: Indicate by "+" or "o" whether each statement is "true" or "false", respectively: o_ 1. If a primal LP constraint is slack at the

More information

3. Linear Programming and Polyhedral Combinatorics

3. Linear Programming and Polyhedral Combinatorics Massachusetts Institute of Technology Handout 6 18.433: Combinatorial Optimization February 20th, 2009 Michel X. Goemans 3. Linear Programming and Polyhedral Combinatorics Summary of what was seen in the

More information

Duality of linear conic problems

Duality of linear conic problems Duality of linear conic problems Alexander Shapiro and Arkadi Nemirovski Abstract It is well known that the optimal values of a linear programming problem and its dual are equal to each other if at least

More information

160 CHAPTER 4. VECTOR SPACES

160 CHAPTER 4. VECTOR SPACES 160 CHAPTER 4. VECTOR SPACES 4. Rank and Nullity In this section, we look at relationships between the row space, column space, null space of a matrix and its transpose. We will derive fundamental results

More information

Linear Programming Notes V Problem Transformations

Linear Programming Notes V Problem Transformations Linear Programming Notes V Problem Transformations 1 Introduction Any linear programming problem can be rewritten in either of two standard forms. In the first form, the objective is to maximize, the material

More information

Solving Linear Programs

Solving Linear Programs Solving Linear Programs 2 In this chapter, we present a systematic procedure for solving linear programs. This procedure, called the simplex method, proceeds by moving from one feasible solution to another,

More information

Linear Programming. April 12, 2005

Linear Programming. April 12, 2005 Linear Programming April 1, 005 Parts of this were adapted from Chapter 9 of i Introduction to Algorithms (Second Edition) /i by Cormen, Leiserson, Rivest and Stein. 1 What is linear programming? The first

More information

Chapter 6: Sensitivity Analysis

Chapter 6: Sensitivity Analysis Chapter 6: Sensitivity Analysis Suppose that you have just completed a linear programming solution which will have a major impact on your company, such as determining how much to increase the overall production

More information

Support Vector Machines

Support Vector Machines Support Vector Machines Charlie Frogner 1 MIT 2011 1 Slides mostly stolen from Ryan Rifkin (Google). Plan Regularization derivation of SVMs. Analyzing the SVM problem: optimization, duality. Geometric

More information

Linear Programming for Optimization. Mark A. Schulze, Ph.D. Perceptive Scientific Instruments, Inc.

Linear Programming for Optimization. Mark A. Schulze, Ph.D. Perceptive Scientific Instruments, Inc. 1. Introduction Linear Programming for Optimization Mark A. Schulze, Ph.D. Perceptive Scientific Instruments, Inc. 1.1 Definition Linear programming is the name of a branch of applied mathematics that

More information

Two-Stage Stochastic Linear Programs

Two-Stage Stochastic Linear Programs Two-Stage Stochastic Linear Programs Operations Research Anthony Papavasiliou 1 / 27 Two-Stage Stochastic Linear Programs 1 Short Reviews Probability Spaces and Random Variables Convex Analysis 2 Deterministic

More information

DUAL METHODS IN MIXED INTEGER LINEAR PROGRAMMING

DUAL METHODS IN MIXED INTEGER LINEAR PROGRAMMING DUAL METHODS IN MIXED INTEGER LINEAR PROGRAMMING by Menal Guzelsoy Presented to the Graduate and Research Committee of Lehigh University in Candidacy for the Degree of Doctor of Philosophy in Industrial

More information

Arrangements And Duality

Arrangements And Duality Arrangements And Duality 3.1 Introduction 3 Point configurations are tbe most basic structure we study in computational geometry. But what about configurations of more complicated shapes? For example,

More information

Nonlinear Optimization: Algorithms 3: Interior-point methods

Nonlinear Optimization: Algorithms 3: Interior-point methods Nonlinear Optimization: Algorithms 3: Interior-point methods INSEAD, Spring 2006 Jean-Philippe Vert Ecole des Mines de Paris Jean-Philippe.Vert@mines.org Nonlinear optimization c 2006 Jean-Philippe Vert,

More information

Several Views of Support Vector Machines

Several Views of Support Vector Machines Several Views of Support Vector Machines Ryan M. Rifkin Honda Research Institute USA, Inc. Human Intention Understanding Group 2007 Tikhonov Regularization We are considering algorithms of the form min

More information

Optimization Modeling for Mining Engineers

Optimization Modeling for Mining Engineers Optimization Modeling for Mining Engineers Alexandra M. Newman Division of Economics and Business Slide 1 Colorado School of Mines Seminar Outline Linear Programming Integer Linear Programming Slide 2

More information

Walrasian Demand. u(x) where B(p, w) = {x R n + : p x w}.

Walrasian Demand. u(x) where B(p, w) = {x R n + : p x w}. Walrasian Demand Econ 2100 Fall 2015 Lecture 5, September 16 Outline 1 Walrasian Demand 2 Properties of Walrasian Demand 3 An Optimization Recipe 4 First and Second Order Conditions Definition Walrasian

More information

GENERALIZED INTEGER PROGRAMMING

GENERALIZED INTEGER PROGRAMMING Professor S. S. CHADHA, PhD University of Wisconsin, Eau Claire, USA E-mail: schadha@uwec.edu Professor Veena CHADHA University of Wisconsin, Eau Claire, USA E-mail: chadhav@uwec.edu GENERALIZED INTEGER

More information

Discrete Optimization

Discrete Optimization Discrete Optimization [Chen, Batson, Dang: Applied integer Programming] Chapter 3 and 4.1-4.3 by Johan Högdahl and Victoria Svedberg Seminar 2, 2015-03-31 Todays presentation Chapter 3 Transforms using

More information

26 Linear Programming

26 Linear Programming The greatest flood has the soonest ebb; the sorest tempest the most sudden calm; the hottest love the coldest end; and from the deepest desire oftentimes ensues the deadliest hate. Th extremes of glory

More information

3. Evaluate the objective function at each vertex. Put the vertices into a table: Vertex P=3x+2y (0, 0) 0 min (0, 5) 10 (15, 0) 45 (12, 2) 40 Max

3. Evaluate the objective function at each vertex. Put the vertices into a table: Vertex P=3x+2y (0, 0) 0 min (0, 5) 10 (15, 0) 45 (12, 2) 40 Max SOLUTION OF LINEAR PROGRAMMING PROBLEMS THEOREM 1 If a linear programming problem has a solution, then it must occur at a vertex, or corner point, of the feasible set, S, associated with the problem. Furthermore,

More information

Duality in General Programs. Ryan Tibshirani Convex Optimization 10-725/36-725

Duality in General Programs. Ryan Tibshirani Convex Optimization 10-725/36-725 Duality in General Programs Ryan Tibshirani Convex Optimization 10-725/36-725 1 Last time: duality in linear programs Given c R n, A R m n, b R m, G R r n, h R r : min x R n c T x max u R m, v R r b T

More information

OPTIMIZATION. Schedules. Notation. Index

OPTIMIZATION. Schedules. Notation. Index Easter Term 00 Richard Weber OPTIMIZATION Contents Schedules Notation Index iii iv v Preliminaries. Linear programming............................ Optimization under constraints......................3

More information

Cyber-Security Analysis of State Estimators in Power Systems

Cyber-Security Analysis of State Estimators in Power Systems Cyber-Security Analysis of State Estimators in Electric Power Systems André Teixeira 1, Saurabh Amin 2, Henrik Sandberg 1, Karl H. Johansson 1, and Shankar Sastry 2 ACCESS Linnaeus Centre, KTH-Royal Institute

More information

Solutions to Math 51 First Exam January 29, 2015

Solutions to Math 51 First Exam January 29, 2015 Solutions to Math 5 First Exam January 29, 25. ( points) (a) Complete the following sentence: A set of vectors {v,..., v k } is defined to be linearly dependent if (2 points) there exist c,... c k R, not

More information

Largest Fixed-Aspect, Axis-Aligned Rectangle

Largest Fixed-Aspect, Axis-Aligned Rectangle Largest Fixed-Aspect, Axis-Aligned Rectangle David Eberly Geometric Tools, LLC http://www.geometrictools.com/ Copyright c 1998-2016. All Rights Reserved. Created: February 21, 2004 Last Modified: February

More information

On Minimal Valid Inequalities for Mixed Integer Conic Programs

On Minimal Valid Inequalities for Mixed Integer Conic Programs On Minimal Valid Inequalities for Mixed Integer Conic Programs Fatma Kılınç Karzan June 27, 2013 Abstract We study mixed integer conic sets involving a general regular (closed, convex, full dimensional,

More information

LECTURE: INTRO TO LINEAR PROGRAMMING AND THE SIMPLEX METHOD, KEVIN ROSS MARCH 31, 2005

LECTURE: INTRO TO LINEAR PROGRAMMING AND THE SIMPLEX METHOD, KEVIN ROSS MARCH 31, 2005 LECTURE: INTRO TO LINEAR PROGRAMMING AND THE SIMPLEX METHOD, KEVIN ROSS MARCH 31, 2005 DAVID L. BERNICK dbernick@soe.ucsc.edu 1. Overview Typical Linear Programming problems Standard form and converting

More information

Lecture 2: Homogeneous Coordinates, Lines and Conics

Lecture 2: Homogeneous Coordinates, Lines and Conics Lecture 2: Homogeneous Coordinates, Lines and Conics 1 Homogeneous Coordinates In Lecture 1 we derived the camera equations λx = P X, (1) where x = (x 1, x 2, 1), X = (X 1, X 2, X 3, 1) and P is a 3 4

More information

GI01/M055 Supervised Learning Proximal Methods

GI01/M055 Supervised Learning Proximal Methods GI01/M055 Supervised Learning Proximal Methods Massimiliano Pontil (based on notes by Luca Baldassarre) (UCL) Proximal Methods 1 / 20 Today s Plan Problem setting Convex analysis concepts Proximal operators

More information

No: 10 04. Bilkent University. Monotonic Extension. Farhad Husseinov. Discussion Papers. Department of Economics

No: 10 04. Bilkent University. Monotonic Extension. Farhad Husseinov. Discussion Papers. Department of Economics No: 10 04 Bilkent University Monotonic Extension Farhad Husseinov Discussion Papers Department of Economics The Discussion Papers of the Department of Economics are intended to make the initial results

More information

NP-Hardness Results Related to PPAD

NP-Hardness Results Related to PPAD NP-Hardness Results Related to PPAD Chuangyin Dang Dept. of Manufacturing Engineering & Engineering Management City University of Hong Kong Kowloon, Hong Kong SAR, China E-Mail: mecdang@cityu.edu.hk Yinyu

More information

Recovery of primal solutions from dual subgradient methods for mixed binary linear programming; a branch-and-bound approach

Recovery of primal solutions from dual subgradient methods for mixed binary linear programming; a branch-and-bound approach MASTER S THESIS Recovery of primal solutions from dual subgradient methods for mixed binary linear programming; a branch-and-bound approach PAULINE ALDENVIK MIRJAM SCHIERSCHER Department of Mathematical

More information

Høgskolen i Narvik Sivilingeniørutdanningen STE6237 ELEMENTMETODER. Oppgaver

Høgskolen i Narvik Sivilingeniørutdanningen STE6237 ELEMENTMETODER. Oppgaver Høgskolen i Narvik Sivilingeniørutdanningen STE637 ELEMENTMETODER Oppgaver Klasse: 4.ID, 4.IT Ekstern Professor: Gregory A. Chechkin e-mail: chechkin@mech.math.msu.su Narvik 6 PART I Task. Consider two-point

More information

Optimal shift scheduling with a global service level constraint

Optimal shift scheduling with a global service level constraint Optimal shift scheduling with a global service level constraint Ger Koole & Erik van der Sluis Vrije Universiteit Division of Mathematics and Computer Science De Boelelaan 1081a, 1081 HV Amsterdam The

More information

The Steepest Descent Algorithm for Unconstrained Optimization and a Bisection Line-search Method

The Steepest Descent Algorithm for Unconstrained Optimization and a Bisection Line-search Method The Steepest Descent Algorithm for Unconstrained Optimization and a Bisection Line-search Method Robert M. Freund February, 004 004 Massachusetts Institute of Technology. 1 1 The Algorithm The problem

More information

What is Linear Programming?

What is Linear Programming? Chapter 1 What is Linear Programming? An optimization problem usually has three essential ingredients: a variable vector x consisting of a set of unknowns to be determined, an objective function of x to

More information

15 Kuhn -Tucker conditions

15 Kuhn -Tucker conditions 5 Kuhn -Tucker conditions Consider a version of the consumer problem in which quasilinear utility x 2 + 4 x 2 is maximised subject to x +x 2 =. Mechanically applying the Lagrange multiplier/common slopes

More information

24. The Branch and Bound Method

24. The Branch and Bound Method 24. The Branch and Bound Method It has serious practical consequences if it is known that a combinatorial problem is NP-complete. Then one can conclude according to the present state of science that no

More information

Solving Linear Systems, Continued and The Inverse of a Matrix

Solving Linear Systems, Continued and The Inverse of a Matrix , Continued and The of a Matrix Calculus III Summer 2013, Session II Monday, July 15, 2013 Agenda 1. The rank of a matrix 2. The inverse of a square matrix Gaussian Gaussian solves a linear system by reducing

More information

TOPIC 4: DERIVATIVES

TOPIC 4: DERIVATIVES TOPIC 4: DERIVATIVES 1. The derivative of a function. Differentiation rules 1.1. The slope of a curve. The slope of a curve at a point P is a measure of the steepness of the curve. If Q is a point on the

More information

1 Norms and Vector Spaces

1 Norms and Vector Spaces 008.10.07.01 1 Norms and Vector Spaces Suppose we have a complex vector space V. A norm is a function f : V R which satisfies (i) f(x) 0 for all x V (ii) f(x + y) f(x) + f(y) for all x,y V (iii) f(λx)

More information

Linear Programming Problems

Linear Programming Problems Linear Programming Problems Linear programming problems come up in many applications. In a linear programming problem, we have a function, called the objective function, which depends linearly on a number

More information

Linear Programming. Solving LP Models Using MS Excel, 18

Linear Programming. Solving LP Models Using MS Excel, 18 SUPPLEMENT TO CHAPTER SIX Linear Programming SUPPLEMENT OUTLINE Introduction, 2 Linear Programming Models, 2 Model Formulation, 4 Graphical Linear Programming, 5 Outline of Graphical Procedure, 5 Plotting

More information

MATH2210 Notebook 1 Fall Semester 2016/2017. 1 MATH2210 Notebook 1 3. 1.1 Solving Systems of Linear Equations... 3

MATH2210 Notebook 1 Fall Semester 2016/2017. 1 MATH2210 Notebook 1 3. 1.1 Solving Systems of Linear Equations... 3 MATH0 Notebook Fall Semester 06/07 prepared by Professor Jenny Baglivo c Copyright 009 07 by Jenny A. Baglivo. All Rights Reserved. Contents MATH0 Notebook 3. Solving Systems of Linear Equations........................

More information

The Graphical Method: An Example

The Graphical Method: An Example The Graphical Method: An Example Consider the following linear program: Maximize 4x 1 +3x 2 Subject to: 2x 1 +3x 2 6 (1) 3x 1 +2x 2 3 (2) 2x 2 5 (3) 2x 1 +x 2 4 (4) x 1, x 2 0, where, for ease of reference,

More information

Definition 11.1. Given a graph G on n vertices, we define the following quantities:

Definition 11.1. Given a graph G on n vertices, we define the following quantities: Lecture 11 The Lovász ϑ Function 11.1 Perfect graphs We begin with some background on perfect graphs. graphs. First, we define some quantities on Definition 11.1. Given a graph G on n vertices, we define

More information

Operation Research. Module 1. Module 2. Unit 1. Unit 2. Unit 3. Unit 1

Operation Research. Module 1. Module 2. Unit 1. Unit 2. Unit 3. Unit 1 Operation Research Module 1 Unit 1 1.1 Origin of Operations Research 1.2 Concept and Definition of OR 1.3 Characteristics of OR 1.4 Applications of OR 1.5 Phases of OR Unit 2 2.1 Introduction to Linear

More information

9.4 THE SIMPLEX METHOD: MINIMIZATION

9.4 THE SIMPLEX METHOD: MINIMIZATION SECTION 9 THE SIMPLEX METHOD: MINIMIZATION 59 The accounting firm in Exercise raises its charge for an audit to $5 What number of audits and tax returns will bring in a maximum revenue? In the simplex

More information

Sensitivity Analysis 3.1 AN EXAMPLE FOR ANALYSIS

Sensitivity Analysis 3.1 AN EXAMPLE FOR ANALYSIS Sensitivity Analysis 3 We have already been introduced to sensitivity analysis in Chapter via the geometry of a simple example. We saw that the values of the decision variables and those of the slack and

More information

Introduction to Linear Programming (LP) Mathematical Programming (MP) Concept

Introduction to Linear Programming (LP) Mathematical Programming (MP) Concept Introduction to Linear Programming (LP) Mathematical Programming Concept LP Concept Standard Form Assumptions Consequences of Assumptions Solution Approach Solution Methods Typical Formulations Massachusetts

More information

Solving Systems of Linear Equations

Solving Systems of Linear Equations LECTURE 5 Solving Systems of Linear Equations Recall that we introduced the notion of matrices as a way of standardizing the expression of systems of linear equations In today s lecture I shall show how

More information

[1] Diagonal factorization

[1] Diagonal factorization 8.03 LA.6: Diagonalization and Orthogonal Matrices [ Diagonal factorization [2 Solving systems of first order differential equations [3 Symmetric and Orthonormal Matrices [ Diagonal factorization Recall:

More information

Chapter 6. Cuboids. and. vol(conv(p ))

Chapter 6. Cuboids. and. vol(conv(p )) Chapter 6 Cuboids We have already seen that we can efficiently find the bounding box Q(P ) and an arbitrarily good approximation to the smallest enclosing ball B(P ) of a set P R d. Unfortunately, both

More information

! Solve problem to optimality. ! Solve problem in poly-time. ! Solve arbitrary instances of the problem. #-approximation algorithm.

! Solve problem to optimality. ! Solve problem in poly-time. ! Solve arbitrary instances of the problem. #-approximation algorithm. Approximation Algorithms 11 Approximation Algorithms Q Suppose I need to solve an NP-hard problem What should I do? A Theory says you're unlikely to find a poly-time algorithm Must sacrifice one of three

More information

Linear Programming: Chapter 11 Game Theory

Linear Programming: Chapter 11 Game Theory Linear Programming: Chapter 11 Game Theory Robert J. Vanderbei October 17, 2007 Operations Research and Financial Engineering Princeton University Princeton, NJ 08544 http://www.princeton.edu/ rvdb Rock-Paper-Scissors

More information

Lecture 2: The SVM classifier

Lecture 2: The SVM classifier Lecture 2: The SVM classifier C19 Machine Learning Hilary 2015 A. Zisserman Review of linear classifiers Linear separability Perceptron Support Vector Machine (SVM) classifier Wide margin Cost function

More information

5 INTEGER LINEAR PROGRAMMING (ILP) E. Amaldi Fondamenti di R.O. Politecnico di Milano 1

5 INTEGER LINEAR PROGRAMMING (ILP) E. Amaldi Fondamenti di R.O. Politecnico di Milano 1 5 INTEGER LINEAR PROGRAMMING (ILP) E. Amaldi Fondamenti di R.O. Politecnico di Milano 1 General Integer Linear Program: (ILP) min c T x Ax b x 0 integer Assumption: A, b integer The integrality condition

More information

Using the Simplex Method to Solve Linear Programming Maximization Problems J. Reeb and S. Leavengood

Using the Simplex Method to Solve Linear Programming Maximization Problems J. Reeb and S. Leavengood PERFORMANCE EXCELLENCE IN THE WOOD PRODUCTS INDUSTRY EM 8720-E October 1998 $3.00 Using the Simplex Method to Solve Linear Programming Maximization Problems J. Reeb and S. Leavengood A key problem faced

More information

Minkowski Sum of Polytopes Defined by Their Vertices

Minkowski Sum of Polytopes Defined by Their Vertices Minkowski Sum of Polytopes Defined by Their Vertices Vincent Delos, Denis Teissandier To cite this version: Vincent Delos, Denis Teissandier. Minkowski Sum of Polytopes Defined by Their Vertices. Journal

More information

Dynamic TCP Acknowledgement: Penalizing Long Delays

Dynamic TCP Acknowledgement: Penalizing Long Delays Dynamic TCP Acknowledgement: Penalizing Long Delays Karousatou Christina Network Algorithms June 8, 2010 Karousatou Christina (Network Algorithms) Dynamic TCP Acknowledgement June 8, 2010 1 / 63 Layout

More information

FRACTIONAL COLORINGS AND THE MYCIELSKI GRAPHS

FRACTIONAL COLORINGS AND THE MYCIELSKI GRAPHS FRACTIONAL COLORINGS AND THE MYCIELSKI GRAPHS By G. Tony Jacobs Under the Direction of Dr. John S. Caughman A math 501 project submitted in partial fulfillment of the requirements for the degree of Master

More information

Algorithm Design and Analysis

Algorithm Design and Analysis Algorithm Design and Analysis LECTURE 27 Approximation Algorithms Load Balancing Weighted Vertex Cover Reminder: Fill out SRTEs online Don t forget to click submit Sofya Raskhodnikova 12/6/2011 S. Raskhodnikova;

More information