Sherali-Adams Relaxations of Graph Isomorphism Polytopes. Peter N. Malkin* and Mohammed Omar + UC Davis

Size: px
Start display at page:

Download "Sherali-Adams Relaxations of Graph Isomorphism Polytopes. Peter N. Malkin* and Mohammed Omar + UC Davis"

Transcription

1 Sherali-Adams Relaxations of Graph Isomorphism Polytopes Peter N. Malkin* and Mohammed Omar + UC Davis University of Washington May 18th 2010 *Partly funded by an IBM OCR grant and the NSF. + Partly funded by NSERC Postgraduate Scholarship.

2 Overview There has recently been much interest in project & lift (L&P) approaches that construct hierarchies of LP or SDP relaxations of 0-1 problems: Techniques: Sherali-Adams (1990), Lovász and Schrijver (1991) and Lasserre (2001) among others. Problems: Max-Cut, Stable Set and Vertex Cover for example. We are interested in applying the Sherali-Adams (S-A) technique to the Graph Isomorphism (ISO) and Graph Automorphism (AUT) problems. It is not known whether ISO and AUT are in P or NPC. Our Goal! Find complexity results about L&P applied to ISO & AUT that will hopefully shed light on L&P and the complexity of ISO and AUT.

3 Talk Outline Overview: 1 Combinatorial approaches to ISO and AUT. 2 Polyhedral approaches (Sherali-Adams). 3 Comparison of the two approaches. Notation: We will always use V := {1,.., n} as the vertex set and use simple undirected graphs G = (V, E G ) and H = (V, E H ). An isomorphism from G to H is a bijection ψ : V V s.t. {u, v} E G iff {ψ(u), ψ(v)} E H. The set of all isomorphisms from G to H is ISO(G, H). The set of all automorphisms is AUT (G) = ISO(G, G). We say G is asymmetric if AUT (G) = 1.

4 Vertex Classification (V-C) Algorithm The V-C algorithm for AUT (G) partitions V into equivalence classes {V 1, V 2,..., V m } s.t. ψ(v i ) = V i for all ψ AUT (G). If the partition is complete ( V i = 1), then G is asymmetric. Start from the partition {V } and iteratively refine it using the number of neighbors of a vertex w.r.t. each equivalence class.

5 Vertex Classification (V-C) Algorithm The V-C algorithm for AUT (G) partitions V into equivalence classes {V 1, V 2,..., V m } s.t. ψ(v i ) = V i for all ψ AUT (G). If the partition is complete ( V i = 1), then G is asymmetric. Start from the partition {V } and iteratively refine it using the number of neighbors of a vertex w.r.t. each equivalence class.

6 Vertex Classification (V-C) Algorithm The V-C algorithm for AUT (G) partitions V into equivalence classes {V 1, V 2,..., V m } s.t. ψ(v i ) = V i for all ψ AUT (G). If the partition is complete ( V i = 1), then G is asymmetric. Start from the partition {V } and iteratively refine it using the number of neighbors of a vertex w.r.t. each equivalence class.

7 Vertex Classification (V-C) Algorithm The V-C algorithm for AUT (G) partitions V into equivalence classes {V 1, V 2,..., V m } s.t. ψ(v i ) = V i for all ψ AUT (G). If the partition is complete ( V i = 1), then G is asymmetric. Start from the partition {V } and iteratively refine it using the number of neighbors of a vertex w.r.t. each equivalence class.

8 Vertex Classification (V-C) Algorithm The V-C algorithm for AUT (G) partitions V into equivalence classes {V 1, V 2,..., V m } s.t. ψ(v i ) = V i for all ψ AUT (G). If the partition is complete ( V i = 1), then G is asymmetric. Start from the partition {V } and iteratively refine it using the number of neighbors of a vertex w.r.t. each equivalence class.

9 Vertex Classification (V-C) Algorithm for AUT Define δ G (u) := {v V : {u, v} E G }. V-C algorithm for AUT (G) Start from {V }. Given {V 1, V 2,..., V m }, the vertices u, v V i are in the same refined equivalence class in the next iteration iff for all j, δ G (u) V j = δ G (v) V j. Repeat this process until the partition stabilizes. Let π(δ G ) denote this stable partition. If π(δ G ) is complete, then return YES (asymmetric), otherwise return MAYBE and output π(δ G ).

10 Vertex Classification (V-C) Algorithm for ISO The V-C method for ISO is based upon the V-C method for AUT. First, we modify the V-C method for AUT (G) to return an ordered partition π(δ G ) s.t. the order of the sets in the partition is invariant under isomorphism. Let π(δ G ) = (V 1,..., V mg ) and π(δ H ) = (W 1,..., W mh ). If G H, then m G = m H and ψ(v i ) = W i ψ ISO(G, H). We say π(δ G ) matches π(δ H ) if m G = m H, V i = W i and δ G (u) V j = δ H (v) W j for all i, j and u V i, v W i. The V-C algorithm for ISO(G, H) Compute π(δ G ) := (V 1,..., V mg ) and π(δ H ) := (W 1,..., W mh ). If π(δ G ) does not match π(δ H ) return NO, otherwise return MAYBE and output π(δ G ) and π(δ H ).

11 Vertex Classification (V-C) Algorithm V-C is well-known (by Corneil and Gotleib 1970??). V-C is the basis of many implementations of AUT and ISO including the well-known nauty package of McKay. The V-C results after only two refinement steps in a canonical form (CF ) of random graphs with probability 1 e cn (Babai & Kucera 1979). CF (G) G s.t. CF (G) = CF (H) iff G H. E.g., relabel vertices in G in the order they appear in π(δ G ). The partitions π(δ G ) and π(δ H ) can be used with branching. An additional depth-first search yields a CF of all graphs in linear average time (Babai & Kucera 1979). The V-C method is useless for regular graphs.

12 Polyhedral Approach for ISO and AUT Tinhofer (1986) examined the following polyhedron where A G and A H are the adjacency matrices of G and H respectively: P G,H := {X [0, 1] n n : XA G = A H X, Xe = X T e = e}. The integer points in P G,H are permutation matrices in bijection with ISO(G, H), that is, X P G,H {0, 1} n n iff ψ ISO(G, H) s.t. if ψ(u) = v, then X uv = 1, else X uv = 0. Tinhofer 1986 Let π(δ G ) = (V 1,..., V m ) and π(δ H ) = (W 1,..., W m ). 1 P G,H = iff π(δ G ) doesn t match π(δ H ) 2 P G = {I } iff π(δ G ) is complete. 3 If π(δ G ) matches π(δ H ), then X uv = 0 for all X P G,H iff u V i, v W j where i j.

13 k-dim Vertex Classification (V-C) Algorithm Define the set of k-tuples V k := {(u 1,..., u k ) : u 1,..., u k V }. The k-dim V-C algorithm for AUT (G) partitions V k into equivalence classes {V1 k, V 2 k,..., V m} k s.t. ψ(vi k ) = Vi k for all ψ AUT (G) where ψ(u) = (ψ(u 1 ), ψ(u 2 ),..., ψ(u k )). If the partition is complete ( V k i = 1), then G is asymmetric. For the graph below, the partition returned by the 2-dim V-C is {(1, 1), (3, 3)}, {(2, 2), (4, 4)}, {(1, 3), (3, 1)}, {(2, 4), (4, 2)}, {(1, 2), (1, 4),(3, 2), (3, 4)}, {(2, 1), (4, 1),(2, 3), (4, 3)}

14 k-dim Vertex Classification (V-C) Algorithm For the k-dim V-C method, we start from the subgraph partition of V k w.r.t. G and iteratively refine it using the number of neighbors of a k-tuple w.r.t. each equivalence class. The subgraph partition w.r.t. G is πg k where u v iff u s = u t v s = v t and {u s, u t } E G {v s, v t } E G. E.g., the subgraph partition of V 2 w.r.t. G is {V1 2, V 2 2, V 3 2}: 1 V1 2 = {(u, u) : u V } (vertices), 2 V2 2 = {(u 1, u 2 ) : {u 1, u 2 } E G } (edges) and 3 V3 2 = {(u 1, u 2 ) : {u 1, u 2 } E G } (non-edges). 1 2 {(1, 1), (3, 3), (2, 2), (4, 4)}, {(1, 3), (3, 1), (1, 2), (2, 1), (1, 4), 4 3 (4, 1), (2, 3), (3, 2), (3, 4), (4, 3)}, {(2, 4), (4, 2)}

15 k-dim Vertex Classification (V-C) Algorithm We need to extend the concept of neighbor for k-tuples. Define δ i G (u) := {(u 1,..., u i 1, v, u i+1,..., u k ) : v δ G (u i )} and δ i G (u) := {(u 1,..., u i 1, v, u i+1,..., u k ) : v δ G (u i )} For example, δg 2 ((1, 1)) = {(1, 2), (1, 3), (1, 4)}, δg 2 ((2, 2)) = {(2, 1), (2, 3)}, δg 2 ((1, 2)) = {(1, 1), (1, 3)}, δg 2 ((1, 3)) = {(1, 1), (1, 2), (1, 4)}.

16 k-dim Vertex Classification (V-C) Algorithm k-dim V-C algorithm Start from the subgraph partition. Given {V1 k,..., V m}, k the tuples u, v Vi k are in the same refined equivalence class in the next iteration if for all i, j, δ i G (u) V k j = δ i G (v) V k j and δ i G (u) V k j = δ i G (v) V k j. This process is repeated until the partition stabilizes. Let π k (δ G ) denote this stable partition. If π k (δ G ) is complete, then return YES (asymmetric), otherwise return MAYBE and output π k (δ G ). For k = 1, this is the same as the V-C algorithm. For k = n, the algorithm is trivially exact. For ISO(G, H), we must check if π k (δ G ) matches π k (δ H ).

17 k-dim Weisfeiler-Lehman (W-L) Algorithm The k-dim W-L method (k 2) also partitions V k (1968). Define i (u) := {(u 1,..., u i 1, v, u i+1,..., u k ) : v V }. k-dim W-L algorithm for AUT (G) Start from the subgraph partition. Given {V1 k,..., V m}, k the vertices u, v Vs k are in the same refined equivalence class in the next iteration if for all i, j, i (u) V k j = i (v) V k j. This process is repeated until the partition stabilizes. Let π k ( G ) denote this stable partition. If π k ( G ) is complete, then return YES (asymmetric), otherwise return MAYBE and output π k ( G ). For ISO(G, H), we must check if π k ( G ) matches π k ( H ).

18 The k-dim V-C method vs the k-dim W-L method Theorem: Cai, Fürer and Immerman 1992 There is a class of pairs of non-isomorphic graphs (G n, H n ) having O(n) vertices such that the Ω(n)-dim W-L algorithm is needed to distinguish G n and H n. k-dim V-C implies k-dim W-L since i (u) = δ i G (u) δ i G (u). Also, (k + 1)-dim W-L implies k-dim V-C. The projection of the partition π k ( G ) onto the first k 1 components gives the partition π k (δ G ). Corollary: M- & Omar There is a class of pairs of non-isomorphic graphs (G n, H n ) having O(n) vertices such that the Ω(n)-dim V-C algorithm is needed to distinguish G n and H n.

19 Sherali-Adams Relaxations The S-A relaxations of a semi-algebraic set P = {x [0, 1] n f 1 (x) = 0,..., f s (x) = 0} [0, 1] n are a hierarchy of polyhedra P 1 P n such that P n = conv(p {0, 1} n ). 1 First, we generate some valid constraints on P: x i f j (x) = 0 j, I, I k 1, i I x i i I (1 x i ) 0 I J, J k. i J\I 2 Next, we replace x 2 i with x i since x 2 i = x i for x {0, 1} n. 3 Then, we linearize the system by replacing each monomial i I x i with a new variable y I giving the polyhedron ˆP k in y. 4 Lastly, we project ˆP k giving P k := {x : x i = y {i}, y ˆP k }.

20 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: X uw X wv = 0 u, v V, w δ G (v) w δ H (u) X uw 1 = 0 u V, w V X wv 1 = 0 v V, w V 0 X uv 1 u, v V.

21 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. X uw X wv = 0 u, v V, w δ G (v) w δ H (u) X uw 1 = 0 u V, w V X wv 1 = 0 v V, w V 0 X uv 1 u, v V.

22 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. X rs X uw (r,s) I w δ H (v) w δ G (u) ( ) X rs X uw (r,s) I w V ) X wv ( X rs (r,s) I w V X uv (u,v) I (u,v) J\I X wv (r,s) I (r,s) I = 0 I V 2, I k 1, u, v V X rs = 0 I V 2, I k 1, u V, X rs = 0 I V 2, I k 1, v V, (1 X uv ) 0 I J V 2, J k.

23 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. 2 Replace X 2 uv with X uv and replace (u,v) I X uv with Y I. X rs X uw (r,s) I w δ H (v) w δ G (u) ( ) X rs X uw (r,s) I w V ) X wv ( X rs (r,s) I w V X uv (u,v) I (u,v) J\I X wv (r,s) I (r,s) I = 0 I V 2, I k 1, u, v V X rs = 0 I V 2, I k 1, u V, X rs = 0 I V 2, I k 1, v V, (1 X uv ) 0 I J V 2, J k.

24 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. 2 Replace X 2 uv with X uv and replace (u,v) I X uv with Y I. Y I {(u,w)} Y I {(w,v)} = 0 I V 2, I k 1, u, v V, w δ H (v) w δ G (u) Y I {(u,w)} Y I = 0 I V 2, I k 1, u V, w V Y I {(w,v)} Y I = 0 I V 2, I k 1, v V, w V ( 1) K\I Y K 0 I J V 2, J k, Y = 1. I K J

25 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. 2 Replace X 2 uv with X uv and replace (u,v) I X uv with Y I. Y I {(u,w)} Y I {(w,v)} = 0 I V 2, I k 1, u, v V, w δ H (v) w δ G (u) Y I {(u,w)} Y I = 0 I V 2, I k 1, u V, w V Y I {(w,v)} Y I = 0 I V 2, I k 1, v V, w V Y I 0 I V 2, I k, Y = 1.

26 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. 2 Replace Xuv 2 with X uv and replace (u,v) I X uv with Y I. 3 Rewrite Y I as Y u,v where I = u, v := {(u i, v i ) : 1 i k}. Y I {(u,w)} Y I {(w,v)} = 0 I V 2, I k 1, u, v V, w δ H (v) w δ G (u) Y I {(u,w)} Y I = 0 I V 2, I k 1, u V, w V Y I {(w,v)} Y I = 0 I V 2, I k 1, v V, w V Y I 0 I V 2, I k, Y = 1.

27 Sherali-Adams Relaxations of P G,H We examine the S-A relaxations of P G,H (XA G = A H X ) as follows: 1 Multiply each equation by (r,s) I X rs for all I V 2. 2 Replace Xuv 2 with X uv and replace (u,v) I X uv with Y I. 3 Rewrite Y I as Y u,v where I = u, v := {(u i, v i ) : 1 i k}. Y u,w Y w,v = 0 u, v V k, 1 i k, w δh i (v) w δg i (u) Y u,w Y u,v \(ui,v i ) = 0 u, v V k, 1 i k, w i (v) Y w,v Y u,v \(ui,v i ) = 0 u, v V k, 1 i k, w i (u) Y u,v 0 u, v V k, Y = 1.

28 Comparison of S-A and V-C Note Y ˆP k G,H {0, 1}n n iff there exists ψ ISO(G, H) s.t. if ψ(u) = v, then Y u,v = 1, else Y u,v = 0. Theorem: M- and Omar 2010 Let π k (δ G ) = (V k 1,..., V k m) and π k (δ H ) = (W k 1,..., W k m). 1 P k G = {I } iff the partition πk (δ G ) is complete. 2 P k G,H = iff πk (δ G ) doesn t match π k (δ H ). 3 If π k (δ G ) matches π k (δ H ), then Y u,v = 0 for all Y ˆP G,H k iff u Vi k, v Wj k where i j. Corollary: M- & Omar 2010 There exists a class of pairs of non-isomorphic graphs (G n, H n ) with O(n) vertices s.t. the S-A relaxations of P G,H need Ω(n) iterations to converge to the convex hull proving non-isomorphism.

29 Sherali-Adams Relaxations of Q G,H Question: Does the W-L method correspond to the S-A relaxations of some polytope? Consider the following semi-algebraic set Q G,H : X u1 v 1 X u2 v 2 = 0 u 1, v 1 V, u 2 δ G (u 1 ), v 2 δ H (v 1 ) X u1 v 1 X u2 v 2 = 0 u 1, v 1 V, u 2 δ G (u 1 ), v 2 δ H (v 1 ) X uw 1 = 0 u V, w V X wv 1 = 0 v V, w V 0 X uv 1 u, v V. The integer points in Q G,H are in bijection with ISO(G, H).

30 Sherali-Adams Relaxations of Q G,H The kth S-A extended relaxation of Q G,H, written ˆQ k G,H, is Y u,v = 0 u, v V k, u i δ G (u j ), v i δ H (v j ) Y u,v = 0 u, v V k, u i δ G (u j ), v i δ H (v j ) Y u,w Y u,v \(ui,v i ) = 0 u, v V k, 1 i k, w i (v) w i (u) Y w,v Y u,v \(ui,v i ) = 0 u, v V k, 1 i k, Y u,v 0 u, v V k, Y = 1.

31 Comparison of S-A and W-L Theorem: M- & Omar 2010 Let π k ( G ) = (V k 1,..., V k m) and π k ( H ) = (W k 1,..., W k m). 1 Q k G = {I } iff πk ( G ) is complete. 2 Q k G,H = iff πk ( G ) doesn t match π k ( H ). 3 If π k ( G ) matches π k ( H ), then Y u,v = 0 for all Y ˆQ G,H k iff u Vi k, v Wj k where i j. Corollary: M- & Omar 2010 There exists a class of pairs of non-isomorphic graphs (G n, H n ) with O(n) vertices s.t. the S-A relaxations of Q G,H need Ω(n) iterations to converge to the convex hull proving non-isomorphism.

32 Conclusion We have shown the correspondence of the combinatorial algorithms V-C and W-L with the Sherali-Adams relaxations of P G,H and Q G,H respectively. We showed negative results about the convergence of the S-A relaxations of the polytopes P G,H and Q G,H.

Tractable Approximations of Graph Isomorphism.

Tractable Approximations of Graph Isomorphism. Tractable Approximations of Graph Isomorphism. Anuj Dawar University of Cambridge Computer Laboratory visiting RWTH Aachen joint work with Bjarki Holm Darmstadt, 5 February 2014 Graph Isomorphism Graph

More information

Recognizing Graph Theoretic Properties with Polynomial Ideals

Recognizing Graph Theoretic Properties with Polynomial Ideals Claremont Colleges Scholarship @ Claremont All HMC Faculty Publications and Research HMC Faculty Scholarship 1-1-2010 Recognizing Graph Theoretic Properties with Polynomial Ideals Jesus A. De Loera University

More information

Proof Assistants for Graph Non-isomorphism

Proof Assistants for Graph Non-isomorphism Proof Assistants for Graph Non-isomorphism Arjeh M. Cohen 8 January 2007 second lecture of Three aspects of exact computation a tutorial at Mathematics: Algorithms and Proofs (MAP) Leiden, January 8 12,

More information

Elementary Equivalence in Finite Structures

Elementary Equivalence in Finite Structures Elementary Equivalence in Finite Structures Anuj Dawar University of Cambridge Computer Laboratory YuriFest, Berlin, 11 September 2015 When I First Met Yuri When I was a graduate student, I sent Yuri a

More information

Lift-and-Project Techniques and SDP Hierarchies

Lift-and-Project Techniques and SDP Hierarchies MFO seminar on Semidefinite Programming May 30, 2010 Typical combinatorial optimization problem: max c T x s.t. Ax b, x {0, 1} n P := {x R n Ax b} P I := conv(k {0, 1} n ) LP relaxation Integral polytope

More information

Semidefinite and Second Order Cone Programming Seminar Fall 2001 Lecture 5

Semidefinite and Second Order Cone Programming Seminar Fall 2001 Lecture 5 Semidefinite and Second Order Cone Programming Seminar Fall 2001 Lecture 5 Instructor: Farid Alizadeh Scribe: Anton Riabov 10/08/2001 1 Overview We continue studying the maximum eigenvalue SDP, and generalize

More information

Finite Model Theory and Graph Isomorphism. II.

Finite Model Theory and Graph Isomorphism. II. Finite Model Theory and Graph Isomorphism. II. Anuj Dawar University of Cambridge Computer Laboratory visiting RWTH Aachen Beroun, 13 December 2013 Recapitulation Finite Model Theory aims to study the

More information

Notes on the decomposition result of Karlin et al. [2] for the hierarchy of Lasserre by M. Laurent, December 13, 2012

Notes on the decomposition result of Karlin et al. [2] for the hierarchy of Lasserre by M. Laurent, December 13, 2012 Notes on the decomposition result of Karlin et al. [2] for the hierarchy of Lasserre by M. Laurent, December 13, 2012 We present the decomposition result of Karlin et al. [2] for the hierarchy of Lasserre

More information

MINI-TUTORIAL ON SEMI-ALGEBRAIC PROOF SYSTEMS. Albert Atserias Universitat Politècnica de Catalunya Barcelona

MINI-TUTORIAL ON SEMI-ALGEBRAIC PROOF SYSTEMS. Albert Atserias Universitat Politècnica de Catalunya Barcelona MINI-TUTORIAL ON SEMI-ALGEBRAIC PROOF SYSTEMS Albert Atserias Universitat Politècnica de Catalunya Barcelona Part I CONVEX POLYTOPES Convex polytopes as linear inequalities Polytope: P = {x R n : Ax b}

More information

Sherali-Adams relaxations of the matching polytope

Sherali-Adams relaxations of the matching polytope Sherali-Adams relaxations of the matching polytope Claire Mathieu Alistair Sinclair November 17, 2008 Abstract We study the Sherali-Adams lift-and-project hierarchy of linear programming relaxations of

More information

Cutting planes from extended LP formulations

Cutting planes from extended LP formulations Cutting planes from extended LP formulations Merve Bodur University of Wisconsin-Madison mbodur@wisc.edu Sanjeeb Dash IBM Research sanjeebd@us.ibm.com March 7, 2016 Oktay Günlük IBM Research gunluk@us.ibm.com

More information

CS675: Convex and Combinatorial Optimization Fall 2014 Combinatorial Problems as Linear Programs. Instructor: Shaddin Dughmi

CS675: Convex and Combinatorial Optimization Fall 2014 Combinatorial Problems as Linear Programs. Instructor: Shaddin Dughmi CS675: Convex and Combinatorial Optimization Fall 2014 Combinatorial Problems as Linear Programs Instructor: Shaddin Dughmi Outline 1 Introduction 2 Shortest Path 3 Algorithms for Single-Source Shortest

More information

On the matrix-cut rank of polyhedra

On the matrix-cut rank of polyhedra On the matrix-cut rank of polyhedra William Cook and Sanjeeb Dash Computational and Applied Mathematics Rice University August 5, 00 Abstract Lovász and Schrijver (99) described a semi-definite operator

More information

CS675: Convex and Combinatorial Optimization Fall 2016 Combinatorial Problems as Linear and Convex Programs. Instructor: Shaddin Dughmi

CS675: Convex and Combinatorial Optimization Fall 2016 Combinatorial Problems as Linear and Convex Programs. Instructor: Shaddin Dughmi CS675: Convex and Combinatorial Optimization Fall 2016 Combinatorial Problems as Linear and Convex Programs Instructor: Shaddin Dughmi Outline 1 Introduction 2 Shortest Path 3 Algorithms for Single-Source

More information

1 Integer Decomposition Property

1 Integer Decomposition Property CS 598CSC: Combinatorial Optimization Lecture date: Feb 2, 2010 Instructor: Chandra Chekuri Scribe: Siva Theja Maguluri Material taken mostly from [1] (Chapter 19). 1 Integer Decomposition Property A polyhedron

More information

LOW DEGREE NULLSTELLENSATZ CERTIFICATES FOR 3-COLORABILITY

LOW DEGREE NULLSTELLENSATZ CERTIFICATES FOR 3-COLORABILITY LOW DEGREE NULLSTELLENSATZ CERTIFICATES FOR 3-COLORABILITY BO LI, BENJAMIN LOWENSTEIN, AND MOHAMED OMAR Abstract. In a seminal paper, De Loera et. al introduce the algorithm NulLA (Nullstellensatz Linear

More information

Integrality Gaps for Sherali Adams Relaxations

Integrality Gaps for Sherali Adams Relaxations Integrality Gaps for Sherali Adams Relaxations Moses Charikar Princeton University Konstantin Makarychev IBM T.J. Watson Research Center Yury Makarychev Microsoft Research Abstract We prove strong lower

More information

Combinatorial Types of Tropical Eigenvector

Combinatorial Types of Tropical Eigenvector Combinatorial Types of Tropical Eigenvector arxiv:1105.55504 Ngoc Mai Tran Department of Statistics, UC Berkeley Joint work with Bernd Sturmfels 2 / 13 Tropical eigenvalues and eigenvectors Max-plus: (R,,

More information

Affine Geometry and the Discrete Legendre Transfrom

Affine Geometry and the Discrete Legendre Transfrom Affine Geometry and the Discrete Legendre Transfrom Andrey Novoseltsev Department of Mathematics University of Washington April 25, 2008 Andrey Novoseltsev (UW) Affine Geometry April 25, 2008 1 / 24 Outline

More information

WHEN DOES THE POSITIVE SEMIDEFINITENESS CONSTRAINT HELP IN LIFTING PROCEDURES?

WHEN DOES THE POSITIVE SEMIDEFINITENESS CONSTRAINT HELP IN LIFTING PROCEDURES? MATHEMATICS OF OPERATIONS RESEARCH Vol. 6, No. 4, November 00, pp. 796 85 Printed in U.S.A. WHEN DOES THE POSITIVE SEMIDEFINITENESS CONSTRAINT HELP IN LIFTING PROCEDURES? MICHEL X. GOEMANS and LEVENT TUNÇEL

More information

Exploiting Symmetry in Computing Polyhedral Bounds on Network Coding Rate Regions

Exploiting Symmetry in Computing Polyhedral Bounds on Network Coding Rate Regions Exploiting Symmetry in Computing Polyhedral Bounds on Network Coding Rate Regions Jayant Apte John Walsh Department of Electrical and Computer Engineering Drexel University, Philadelphia NetCod, 205 NSF

More information

Cutting Planes for First Level RLT Relaxations of Mixed 0-1 Programs

Cutting Planes for First Level RLT Relaxations of Mixed 0-1 Programs Cutting Planes for First Level RLT Relaxations of Mixed 0-1 Programs 1 Cambridge, July 2013 1 Joint work with Franklin Djeumou Fomeni and Adam N. Letchford Outline 1. Introduction 2. Literature Review

More information

arxiv: v1 [cs.ds] 29 Aug 2015

arxiv: v1 [cs.ds] 29 Aug 2015 APPROXIMATING (UNWEIGHTED) TREE AUGMENTATION VIA LIFT-AND-PROJECT, PART I: STEMLESS TAP JOSEPH CHERIYAN AND ZHIHAN GAO arxiv:1508.07504v1 [cs.ds] 9 Aug 015 Abstract. In Part I, we study a special case

More information

Iterative Algebraic Algorithms for the Recognition of Combinatorial Properties

Iterative Algebraic Algorithms for the Recognition of Combinatorial Properties Iterative Algebraic Algorithms for the Recognition of Combinatorial Properties J.A. De Loera, C. Hillar, P.N. Malkin, M. Omar November 23, 2009 Abstract Many combinatorial optimization problems can be

More information

A notion of Total Dual Integrality for Convex, Semidefinite and Extended Formulations

A 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 information

A Geometric Approach to Graph Isomorphism

A Geometric Approach to Graph Isomorphism A Geometric Approach to Graph Isomorphism Pawan Aurora and Shashank K Mehta Indian Institute of Technology, Kanpur - 208016, India {paurora,skmehta}@cse.iitk.ac.in Abstract. We present an integer linear

More information

Cuts for mixed 0-1 conic programs

Cuts for mixed 0-1 conic programs Cuts for mixed 0-1 conic programs G. Iyengar 1 M. T. Cezik 2 1 IEOR Department Columbia University, New York. 2 GERAD Université de Montréal, Montréal TU-Chemnitz Workshop on Integer Programming and Continuous

More information

Ma/CS 6b Class 12: Graphs and Matrices

Ma/CS 6b Class 12: Graphs and Matrices Ma/CS 6b Class 2: Graphs and Matrices 3 3 v 5 v 4 v By Adam Sheffer Non-simple Graphs In this class we allow graphs to be nonsimple. We allow parallel edges, but not loops. Incidence Matrix Consider a

More information

The Matching Polytope: General graphs

The Matching Polytope: General graphs 8.433 Combinatorial Optimization The Matching Polytope: General graphs September 8 Lecturer: Santosh Vempala A matching M corresponds to a vector x M = (0, 0,,, 0...0) where x M e is iff e M and 0 if e

More information

Graph invariants in the spin model

Graph invariants in the spin model Graph invariants in the spin model Alexander Schrijver 1 Abstract. Given a symmetric n n matrix A, we define, for any graph G, f A (G) := a φ(u),φ(v). φ:v G {1,...,n} uv EG We characterize for which graph

More information

arxiv: v15 [math.oc] 19 Oct 2016

arxiv: v15 [math.oc] 19 Oct 2016 LP formulations for mixed-integer polynomial optimization problems Daniel Bienstock and Gonzalo Muñoz, Columbia University, December 2014 arxiv:1501.00288v15 [math.oc] 19 Oct 2016 Abstract We present a

More information

Combinatorial Commutative Algebra, Graph Colorability, and Algorithms

Combinatorial Commutative Algebra, Graph Colorability, and Algorithms I G,k = g 1,...,g n Combinatorial Commutative Algebra, Graph Colorability, and Algorithms Chris Hillar (University of California, Berkeley) Joint with Troels Windfeldt! (University of Copenhagen)! Outline

More information

3.8 Strong valid inequalities

3.8 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 information

CSC Linear Programming and Combinatorial Optimization Lecture 12: The Lift and Project Method

CSC Linear Programming and Combinatorial Optimization Lecture 12: The Lift and Project Method CSC2411 - Linear Programming and Combinatorial Optimization Lecture 12: The Lift and Project Method Notes taken by Stefan Mathe April 28, 2007 Summary: Throughout the course, we have seen the importance

More information

Introduction to LP and SDP Hierarchies

Introduction to LP and SDP Hierarchies Introduction to LP and SDP Hierarchies Madhur Tulsiani Princeton University Local Constraints in Approximation Algorithms Linear Programming (LP) or Semidefinite Programming (SDP) based approximation algorithms

More information

Lift-and-Project Inequalities

Lift-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 information

Approximation Algorithms for the k-set Packing Problem

Approximation Algorithms for the k-set Packing Problem Approximation Algorithms for the k-set Packing Problem Marek Cygan Institute of Informatics University of Warsaw 20th October 2016, Warszawa Marek Cygan Approximation Algorithms for the k-set Packing Problem

More information

ON THE INTEGRALITY OF THE UNCAPACITATED FACILITY LOCATION POLYTOPE. 1. Introduction

ON THE INTEGRALITY OF THE UNCAPACITATED FACILITY LOCATION POLYTOPE. 1. Introduction ON THE INTEGRALITY OF THE UNCAPACITATED FACILITY LOCATION POLYTOPE MOURAD BAÏOU AND FRANCISCO BARAHONA Abstract We study a system of linear inequalities associated with the uncapacitated facility location

More information

4 Packing T-joins and T-cuts

4 Packing T-joins and T-cuts 4 Packing T-joins and T-cuts Introduction Graft: A graft consists of a connected graph G = (V, E) with a distinguished subset T V where T is even. T-cut: A T -cut of G is an edge-cut C which separates

More information

Fractional coloring. March 24, 2018

Fractional coloring. March 24, 2018 Fractional coloring Zdeněk Dvořák March 24, 2018 1 Introduction and definitions Note that a graph is k-colorable if and only if it can be covered by at most k independent sets; i.e., we can assign value

More information

Applications of semidefinite programming in Algebraic Combinatorics

Applications of semidefinite programming in Algebraic Combinatorics Applications of semidefinite programming in Algebraic Combinatorics Tohoku University The 23rd RAMP Symposium October 24, 2011 We often want to 1 Bound the value of a numerical parameter of certain combinatorial

More information

Some Geometrical and Vertex-Partitioning Techniques for Graph Isomorphism

Some Geometrical and Vertex-Partitioning Techniques for Graph Isomorphism Some Geometrical and Vertex-Partitioning Techniques for Graph Isomorphism By Gaurav Rattan MATH10201005005 The Institute of Mathematical Sciences, Chennai A thesis submitted to the Board of Studies in

More information

Dimension reduction for semidefinite programming

Dimension reduction for semidefinite programming 1 / 22 Dimension reduction for semidefinite programming Pablo A. Parrilo Laboratory for Information and Decision Systems Electrical Engineering and Computer Science Massachusetts Institute of Technology

More information

Lecture Semidefinite Programming and Graph Partitioning

Lecture Semidefinite Programming and Graph Partitioning Approximation Algorithms and Hardness of Approximation April 16, 013 Lecture 14 Lecturer: Alantha Newman Scribes: Marwa El Halabi 1 Semidefinite Programming and Graph Partitioning In previous lectures,

More information

Introduction to Integer Programming

Introduction 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 information

Sherali-Adams Relaxations of the Matching Polytope

Sherali-Adams Relaxations of the Matching Polytope Sherali-Adams Relaxations of the Matching Polytope Claire Mathieu Computer Science Department Brown University Providence, RI 091 claire@cs.brown.edu Alistair Sinclair Computer Science Division University

More information

On the Rank of Cutting-Plane Proof Systems

On the Rank of Cutting-Plane Proof Systems On the Rank of Cutting-Plane Proof Systems Sebastian Pokutta 1 and Andreas S. Schulz 2 1 H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA. Email:

More information

Some Notes on Distance-Transitive and Distance-Regular Graphs

Some Notes on Distance-Transitive and Distance-Regular Graphs Some Notes on Distance-Transitive and Distance-Regular Graphs Robert Bailey February 2004 These are notes from lectures given in the Queen Mary Combinatorics Study Group on 13th and 20th February 2004,

More information

LP formulations for mixed-integer polynomial optimization problems Daniel Bienstock and Gonzalo Muñoz, Columbia University, December 2014

LP formulations for mixed-integer polynomial optimization problems Daniel Bienstock and Gonzalo Muñoz, Columbia University, December 2014 LP formulations for mixed-integer polynomial optimization problems Daniel Bienstock and Gonzalo Muñoz, Columbia University, December 2014 Abstract We present a class of linear programming approximations

More information

3.7 Cutting plane methods

3.7 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 information

The destiny of constant structure discrete time closed semantic systems

The destiny of constant structure discrete time closed semantic systems arxiv:1711.07071v1 [cs.ai] 19 Nov 2017 The destiny of constant structure discrete time closed semantic systems E.E. Ivanko a) Institute of Mathematics and Mechanics, Ural branch, RAS b) Ural Federal University

More information

A NEW SEMIDEFINITE PROGRAMMING HIERARCHY FOR CYCLES IN BINARY MATROIDS AND CUTS IN GRAPHS

A NEW SEMIDEFINITE PROGRAMMING HIERARCHY FOR CYCLES IN BINARY MATROIDS AND CUTS IN GRAPHS A NEW SEMIDEFINITE PROGRAMMING HIERARCHY FOR CYCLES IN BINARY MATROIDS AND CUTS IN GRAPHS JOÃO GOUVEIA, MONIQUE LAURENT, PABLO A. PARRILO, AND REKHA THOMAS Abstract. The theta bodies of a polynomial ideal

More information

Computing autotopism groups of partial Latin rectangles: a pilot study

Computing autotopism groups of partial Latin rectangles: a pilot study Computing autotopism groups of partial Latin rectangles: a pilot study Raúl M. Falcón (U. Seville ); Daniel Kotlar (Tel-Hai College ); Rebecca J. Stones (Nankai U. ) 9 December 206 2 3 2 4 5 6 7 5 3 4

More information

Abstract We survey some of what can be deduced about automorphisms of a graph from information on its eigenvalues and eigenvectors. Two of the main to

Abstract We survey some of what can be deduced about automorphisms of a graph from information on its eigenvalues and eigenvectors. Two of the main to Symmetry and Eigenvectors Ada Chan and Chris D. Godsil Combinatorics and Optimization University of Waterloo Waterloo, Ontario Canada N2L 3G http://bilby.uwaterloo.ca Support from a National Sciences and

More information

Introduction to Linear and Combinatorial Optimization (ADM I)

Introduction to Linear and Combinatorial Optimization (ADM I) Introduction to Linear and Combinatorial Optimization (ADM I) Rolf Möhring based on the 20011/12 course by Martin Skutella TU Berlin WS 2013/14 1 General Remarks new flavor of ADM I introduce linear and

More information

The Graph Isomorphism Problem and the Module Isomorphism Problem. Harm Derksen

The Graph Isomorphism Problem and the Module Isomorphism Problem. Harm Derksen The Graph Isomorphism Problem and the Module Isomorphism Problem Harm Derksen Department of Mathematics Michigan Center for Integrative Research in Critical Care University of Michigan Partially supported

More information

USING GRAPHS AND GAMES TO GENERATE CAP SET BOUNDS

USING GRAPHS AND GAMES TO GENERATE CAP SET BOUNDS USING GRAPHS AND GAMES TO GENERATE CAP SET BOUNDS JOSH ABBOTT AND TREVOR MCGUIRE Abstract. Let F 3 be the field with 3 elements and consider the k- dimensional affine space, F k 3, over F 3. A line of

More information

Deciding Emptiness of the Gomory-Chvátal Closure is NP-Complete, Even for a Rational Polyhedron Containing No Integer Point

Deciding Emptiness of the Gomory-Chvátal Closure is NP-Complete, Even for a Rational Polyhedron Containing No Integer Point Deciding Emptiness of the Gomory-Chvátal Closure is NP-Complete, Even for a Rational Polyhedron Containing No Integer Point Gérard Cornuéjols 1 and Yanjun Li 2 1 Tepper School of Business, Carnegie Mellon

More information

Integer programming: an introduction. Alessandro Astolfi

Integer programming: an introduction. Alessandro Astolfi Integer programming: an introduction Alessandro Astolfi Outline Introduction Examples Methods for solving ILP Optimization on graphs LP problems with integer solutions Summary Introduction Integer programming

More information

LOVÁSZ-SCHRIJVER SDP-OPERATOR AND A SUPERCLASS OF NEAR-PERFECT GRAPHS

LOVÁSZ-SCHRIJVER SDP-OPERATOR AND A SUPERCLASS OF NEAR-PERFECT GRAPHS LOVÁSZ-SCHRIJVER SDP-OPERATOR AND A SUPERCLASS OF NEAR-PERFECT GRAPHS S. BIANCHI, M. ESCALANTE, G. NASINI, L. TUNÇEL Abstract. We study the Lovász-Schrijver SDP-operator applied to the fractional stable

More information

- Well-characterized problems, min-max relations, approximate certificates. - LP problems in the standard form, primal and dual linear programs

- 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

Hierarchies. 1. Lovasz-Schrijver (LS), LS+ 2. Sherali Adams 3. Lasserre 4. Mixed Hierarchy (recently used) Idea: P = conv(subset S of 0,1 n )

Hierarchies. 1. Lovasz-Schrijver (LS), LS+ 2. Sherali Adams 3. Lasserre 4. Mixed Hierarchy (recently used) Idea: P = conv(subset S of 0,1 n ) Hierarchies Today 1. Some more familiarity with Hierarchies 2. Examples of some basic upper and lower bounds 3. Survey of recent results (possible material for future talks) Hierarchies 1. Lovasz-Schrijver

More information

Artin-Schelter regular algebras and the Steenrod algebra

Artin-Schelter regular algebras and the Steenrod algebra Artin-Schelter regular algebras and the Steenrod algebra J. H. Palmieri and J. J. Zhang University of Washington Los Angeles, 10 October 2010 Exercise Let A(1) be the sub-hopf algebra of the mod 2 Steenrod

More information

DISTINGUISHING PARTITIONS AND ASYMMETRIC UNIFORM HYPERGRAPHS

DISTINGUISHING PARTITIONS AND ASYMMETRIC UNIFORM HYPERGRAPHS DISTINGUISHING PARTITIONS AND ASYMMETRIC UNIFORM HYPERGRAPHS M. N. ELLINGHAM AND JUSTIN Z. SCHROEDER In memory of Mike Albertson. Abstract. A distinguishing partition for an action of a group Γ on a set

More information

RECAP: Extremal problems Examples

RECAP: Extremal problems Examples RECAP: Extremal problems Examples Proposition 1. If G is an n-vertex graph with at most n edges then G is disconnected. A Question you always have to ask: Can we improve on this proposition? Answer. NO!

More information

linear programming and approximate constraint satisfaction

linear programming and approximate constraint satisfaction linear programming and approximate constraint satisfaction Siu On Chan MSR New England James R. Lee University of Washington Prasad Raghavendra UC Berkeley David Steurer Cornell results MAIN THEOREM: Any

More information

Moments and Positive Polynomials for Optimization II: LP- VERSUS SDP-relaxations

Moments and Positive Polynomials for Optimization II: LP- VERSUS SDP-relaxations Moments and Positive Polynomials for Optimization II: LP- VERSUS SDP-relaxations LAAS-CNRS and Institute of Mathematics, Toulouse, France EECI Course: February 2016 LP-relaxations LP- VERSUS SDP-relaxations

More information

Integer Programming Methods LNMB

Integer Programming Methods LNMB Integer Programming Methods LNMB 2017 2018 Dion Gijswijt homepage.tudelft.nl/64a8q/intpm/ Dion Gijswijt Intro IntPM 2017-2018 1 / 24 Organisation Webpage: homepage.tudelft.nl/64a8q/intpm/ Book: Integer

More information

Optimization methods NOPT048

Optimization methods NOPT048 Optimization methods NOPT048 Jirka Fink https://ktiml.mff.cuni.cz/ fink/ Department of Theoretical Computer Science and Mathematical Logic Faculty of Mathematics and Physics Charles University in Prague

More information

Lecture 22: Counting

Lecture 22: Counting CS 710: Complexity Theory 4/8/2010 Lecture 22: Counting Instructor: Dieter van Melkebeek Scribe: Phil Rydzewski & Chi Man Liu Last time we introduced extractors and discussed two methods to construct them.

More information

Semidefinite Programming and Harmonic Analysis

Semidefinite Programming and Harmonic Analysis 1 / 74 Semidefinite Programming and Harmonic Analysis Cristóbal Guzmán CS 8803 - Discrete Fourier Analysis and Applications March 7, 2012 2 / 74 Motivation SDP gives best relaxations known for some combinatorial

More information

Cutting Plane Methods I

Cutting Plane Methods I 6.859/15.083 Integer Programming and Combinatorial Optimization Fall 2009 Cutting Planes Consider max{wx : Ax b, x integer}. Cutting Plane Methods I Establishing the optimality of a solution is equivalent

More information

LIFTS OF CONVEX SETS AND CONE FACTORIZATIONS JOÃO GOUVEIA, PABLO A. PARRILO, AND REKHA THOMAS

LIFTS OF CONVEX SETS AND CONE FACTORIZATIONS JOÃO GOUVEIA, PABLO A. PARRILO, AND REKHA THOMAS LIFTS OF CONVEX SETS AND CONE FACTORIZATIONS Abstract. In this paper we address the basic geometric question of when a given convex set is the image under a linear map of an affine slice of a given closed

More information

Edge colored complete bipartite graphs with trivial automorphism groups

Edge colored complete bipartite graphs with trivial automorphism groups Edge colored complete bipartite graphs with trivial automorphism groups Michael J. Fisher Garth Isaak Abstract We determine the values of s and t for which there is a coloring of the edges of the complete

More information

Submodular Functions, Optimization, and Applications to Machine Learning

Submodular Functions, Optimization, and Applications to Machine Learning Submodular Functions, Optimization, and Applications to Machine Learning Spring Quarter, Lecture 12 http://www.ee.washington.edu/people/faculty/bilmes/classes/ee596b_spring_2016/ Prof. Jeff Bilmes University

More information

The Strong Largeur d Arborescence

The Strong Largeur d Arborescence The Strong Largeur d Arborescence Rik Steenkamp (5887321) November 12, 2013 Master Thesis Supervisor: prof.dr. Monique Laurent Local Supervisor: prof.dr. Alexander Schrijver KdV Institute for Mathematics

More information

3.7 Strong valid inequalities for structured ILP problems

3.7 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 information

Relaxations of combinatorial problems via association schemes

Relaxations of combinatorial problems via association schemes 1 Relaxations of combinatorial problems via association schemes Etienne de Klerk, Fernando M. de Oliveira Filho, and Dmitrii V. Pasechnik Tilburg University, The Netherlands; Nanyang Technological University,

More information

AUTOMORPHISM GROUPS AND SPECTRA OF CIRCULANT GRAPHS

AUTOMORPHISM GROUPS AND SPECTRA OF CIRCULANT GRAPHS AUTOMORPHISM GROUPS AND SPECTRA OF CIRCULANT GRAPHS MAX GOLDBERG Abstract. We explore ways to concisely describe circulant graphs, highly symmetric graphs with properties that are easier to generalize

More information

The rank of connection matrices and the dimension of graph algebras

The rank of connection matrices and the dimension of graph algebras The rank of connection matrices and the dimension of graph algebras László Lovász Microsoft Research One Microsoft Way Redmond, WA 98052 August 2004 Microsoft Research Technical Report TR-2004-82 Contents

More information

Change in the State of the Art of (Mixed) Integer Programming

Change in the State of the Art of (Mixed) Integer Programming Change in the State of the Art of (Mixed) Integer Programming 1977 Vancouver Advanced Research Institute 24 papers 16 reports 1 paper computational, 4 small instances Report on computational aspects: only

More information

arxiv:math/ v1 [math.co] 23 May 2000

arxiv:math/ v1 [math.co] 23 May 2000 Some Fundamental Properties of Successive Convex Relaxation Methods on LCP and Related Problems arxiv:math/0005229v1 [math.co] 23 May 2000 Masakazu Kojima Department of Mathematical and Computing Sciences

More information

Orbitopes. Marc Pfetsch. joint work with Volker Kaibel. Zuse Institute Berlin

Orbitopes. Marc Pfetsch. joint work with Volker Kaibel. Zuse Institute Berlin Orbitopes Marc Pfetsch joint work with Volker Kaibel Zuse Institute Berlin What this talk is about We introduce orbitopes. A polyhedral way to break symmetries in integer programs. Introduction 2 Orbitopes

More information

On mixed-integer sets with two integer variables

On mixed-integer sets with two integer variables On mixed-integer sets with two integer variables Sanjeeb Dash IBM Research sanjeebd@us.ibm.com Santanu S. Dey Georgia Inst. Tech. santanu.dey@isye.gatech.edu September 8, 2010 Oktay Günlük IBM Research

More information

MA441: Algebraic Structures I. Lecture 26

MA441: Algebraic Structures I. Lecture 26 MA441: Algebraic Structures I Lecture 26 10 December 2003 1 (page 179) Example 13: A 4 has no subgroup of order 6. BWOC, suppose H < A 4 has order 6. Then H A 4, since it has index 2. Thus A 4 /H has order

More information

arxiv: v2 [cs.ds] 30 Apr 2014

arxiv: v2 [cs.ds] 30 Apr 2014 Dimension Reduction via Colour Refinement Martin Grohe * RWTH Aachen University grohe@informatik.rwth-aachen.de Martin Mladenov TU Dortmund University martin.mladenov@cs.tu-dortmund.de Kristian Kersting

More information

R(p, k) = the near regular complete k-partite graph of order p. Let p = sk+r, where s and r are positive integers such that 0 r < k.

R(p, k) = the near regular complete k-partite graph of order p. Let p = sk+r, where s and r are positive integers such that 0 r < k. MATH3301 EXTREMAL GRAPH THEORY Definition: A near regular complete multipartite graph is a complete multipartite graph with orders of its partite sets differing by at most 1. R(p, k) = the near regular

More information

Moments and Positive Polynomials for Optimization II: LP- VERSUS SDP-relaxations

Moments and Positive Polynomials for Optimization II: LP- VERSUS SDP-relaxations Moments and Positive Polynomials for Optimization II: LP- VERSUS SDP-relaxations LAAS-CNRS and Institute of Mathematics, Toulouse, France Tutorial, IMS, Singapore 2012 LP-relaxations LP- VERSUS SDP-relaxations

More information

A short course on matching theory, ECNU Shanghai, July 2011.

A short course on matching theory, ECNU Shanghai, July 2011. A short course on matching theory, ECNU Shanghai, July 2011. Sergey Norin LECTURE 3 Tight cuts, bricks and braces. 3.1. Outline of Lecture Ear decomposition of bipartite graphs. Tight cut decomposition.

More information

Spectral Graph Theory

Spectral Graph Theory Spectral Graph Theory Aaron Mishtal April 27, 2016 1 / 36 Outline Overview Linear Algebra Primer History Theory Applications Open Problems Homework Problems References 2 / 36 Outline Overview Linear Algebra

More information

Partial cubes: structures, characterizations, and constructions

Partial cubes: structures, characterizations, and constructions Partial cubes: structures, characterizations, and constructions Sergei Ovchinnikov San Francisco State University, Mathematics Department, 1600 Holloway Ave., San Francisco, CA 94132 Abstract Partial cubes

More information

An algebraic proof of the Erdős-Ko-Rado theorem for intersecting families of perfect matchings

An algebraic proof of the Erdős-Ko-Rado theorem for intersecting families of perfect matchings Also available at http://amc-journal.eu ISSN 855-3966 (printed edn.), ISSN 855-3974 (electronic edn.) ARS MATHEMATICA CONTEMPORANEA 2 (207) 205 27 An algebraic proof of the Erdős-Ko-Rado theorem for intersecting

More information

Symmetry in Network Coding

Symmetry in Network Coding Symmetry in Network Coding Formalization, Graph-theoretic Characterization, and Computation Jayant Apte John Walsh Department of Electrical and Computer Engineering Drexel University, Philadelphia ISIT,

More information

On highly closed cellular algebras and highly closed isomorphisms

On highly closed cellular algebras and highly closed isomorphisms 1 On highly closed cellular algebras and highly closed isomorphisms Dedicated to A. A. Lehman and B. Yu. Weisfeiler on the occasion of the 30th anniversary of their paper where the cellular algebra first

More information

BBM402-Lecture 20: LP Duality

BBM402-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 information

Abstract Algebra II Groups ( )

Abstract Algebra II Groups ( ) Abstract Algebra II Groups ( ) Melchior Grützmann / melchiorgfreehostingcom/algebra October 15, 2012 Outline Group homomorphisms Free groups, free products, and presentations Free products ( ) Definition

More information

CO 250 Final Exam Guide

CO 250 Final Exam Guide Spring 2017 CO 250 Final Exam Guide TABLE OF CONTENTS richardwu.ca CO 250 Final Exam Guide Introduction to Optimization Kanstantsin Pashkovich Spring 2017 University of Waterloo Last Revision: March 4,

More information

Not Every GTSP Facet Induces an STSP Facet

Not Every GTSP Facet Induces an STSP Facet Not Every GTSP Facet Induces an STSP Facet Marcus Oswald, Gerhard Reinelt, and Dirk Oliver Theis Institute of Computer Science University of Heidelberg Im Neuenheimer Feld 38, 9120 Heidelberg, Germany

More information

Cutting Plane Methods II

Cutting Plane Methods II 6.859/5.083 Integer Programming and Combinatorial Optimization Fall 2009 Cutting Plane Methods II Gomory-Chvátal cuts Reminder P = {x R n : Ax b} with A Z m n, b Z m. For λ [0, ) m such that λ A Z n, (λ

More information

Homogeneous Structures and Ramsey Classes

Homogeneous Structures and Ramsey Classes Homogeneous Structures and Ramsey Classes Julia Böttcher April 27, 2005 In this project report we discuss the relation between Ramsey classes and homogeneous structures and give various examples for both

More information