DOCSLIB.ORG

Matching Is As Easy As Matrix Inversion

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Matching Is As Easy As Matrix Inversion Matching is as Easy as Matrix Inversion Ketan Mulmuley ’ Computer Science Department University of California, Berkeley Umesh V. Vazirani 2 Aiken Computation Laboratory Harvard University Vijay V. Vazirani 3 Computer Science Department Cornell University Abstract only computationally non-trivial step required in its execution is the inversion of a A new algorithm for finding a maximum single integer matrix. Since this step can be matching in a general graph is presented; its parallel&d, we get a simple parallel (RNC2) special feature being that the only computa- algorithm. Because of this simplicity, the tionally non-trivial step required in its execu- sequential version of our algorithm has some tion is the inversion of a single integer merits over the conventional matching algo- matrix. Since this step can be parallelized, rithms as well. we get a simple parallel (RNC2) algorithm. This algorithm was obtained while At the heart of our algorithm lies a proba- solving the matching problem from the bilistic lemma, the isolating lemma. We viewpoint of parallel computation. The main show applications of this lemma to parallel difficulty here is that the graph may contain computation and randomized reductions. exponentially many maximum matchings; how do we coordinate the processors so they seek the same matching in parallel? The key to achieving this coordination is a probabilis- 1. Introduction tic lemma, the isolating lemma, which lies at the heart of our algorithm; it helps to single A new algorithm for finding a max- out one matching in the graph. imum matching in a general graph is In its general form, the isolating presented; its special feature being that the lemma holds for an arbitrary set system. This yields a relationship between the paral- Permission to copy without fee all or part of this lei complexity of an arbitrary search problem material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication ’ Miller Fellow, University of California, Berkeley. and its date appear, and notice is given that copying is s Work done during a post-doctoral position at MSRI, by permission of the Association for Computing Berkeley. Machinery. To copy otherwise, or to republish, requires a fee and/or specific permission. 3 Supported by NSF Grant BCR 8503611 and an IBM Faculty Development Award. Currently at AT&T Bell Labs, Murray Hill, NJ 07974. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or to republish, requires a fee and/or specific permission. 0 1987ACM O-89791-221-7/87/0006-0345 75c 345 and the corresponding weighted decision Tutte in 1947 ~Tu], based on the work of problem. As an application, we give an Pfaff on skew-symmetric matrices. It states RNC’ algorithm for the Exact Matching that a graph has a perfect matching iff a cer- problem in general graphs. It is interesting tain matrix of indeterminates, called the to note that this problem is not known to be Tutte matrix,, is non-singular. Motivated by in P. The isolating lemma also yields a sim- an a1gorithmi.c use of this theorem, :Edmonds ple proof for the result in IValVaz] showing [Ed21 studied the complexity of computing the NP-hardness, under randomized reduc- determinants.. He gave a modified Gaussian tions, of instances of SAT having a unique elimination procedure for computing the solution. determinant of an integer matrix in a polyno- Matching was first shown to be in mial number of bit operations, and stated the Random NC @NC3 ) by Karp, Upfal and open problem of efficiently deciding whcthcr Wigderson [KUW 11. Their algorithm also a matrix of indeterminates is non-singular. utilizes matrix operations, and in fact these The first algorithm based on Tuttc’s are some of the most widely used tools for theorem was given by Lovasz l&01]. Using obtaining fast parallel algorithms. Several the fundamlental insight that polynomial problems are known to be NC’ reducible to identities can be efficiently tested by ran- computing the determinant of an integer domization [SC], Lovasz reduced, the deci- matrix. Cook [Co] defines DET to be the sion problem, ‘Does the given graph have a class of all such problems. DET G NC2, and perfect matching? ’ to testing if a given it is not known whether this inclusion is integer matrix is non-singular. Since the proper. All problems known to be in NC2 latter problem is in NC2 [Cs], this yields a are either in AC1 or in DET [Co]. Our algo- (Monte Carlo) RNC2 algorithm for the rithm puts the matching problem in RDET. former problem (see also [BGH]). Rabin and Vazirani IRV] extended this approach, using a theorem of Frobenius, to give a simple ran- domizing algorithm which finds a perfect matching by sequentially inverting IV I /2 2. History matrices. The search problem, i.e. actually find- The maximum matching problem is a ing a perfect matching in parallel, is much natural problem, and its study has led to harder. The first parallel (RNC3) algorithm conceptual breakthroughs in the field of for this long-standing open problem was algorithms. In fact, the characterization of given by Karp, Upfal and Wigderson ‘tractable problems’ as ‘polynomial time solv- [KUWl]. Tbey use the Tutte matrix to able problems’ was first proposed by implement (in RNC2) a ‘rank’ function. Edmonds [Edl] in the context of the general Their algorithm probabilistically prunes out graph matching problem. Solving this prob- edges from the graph; the rank function lem from the viewpoint of parallel computa- guarantees that the remaining graph has a tion has also been quite fruitful. perfect matching, and a probabilistic lemma Whereas sequential algorithms for the ensures that a constant fraction of the edges maximum matching problem are based on are pruned at each stage. Hence, after finding ‘blossoms’ and ‘augmenting paths’ in O(Iog IVI) stages only a perfect matching graphs (see lEdl]), the known parallel algo- remains. This algorithm is also Monte Carlo rithms require a new approach; they use pro- in that it may fail to give a perfect matching. babilistic and algebraic methods. In fact, the Using the Gallai-Edmonds Structure matching problem emerged from algebra Theorem (see [LOO]) Karloff [Ka] gives a around the turn of this century in the works complementary Monte Carlo (RNC2) algo- of Petersen, Frobenius and Konig (for a rithm for bounding the size of a maximum detailed history see [LPI), A key ingredient matching from above, thus yielding a Las in the new approach is a theorem proved by Vegas extension. 346 Our algorithm is conceptually different Since S contains n elements, in that it directly finds a perfect matching in the graph. It is somewhat faster (RI@) and Pr [There exists a singular requires 0 (n3*‘m) processors. element] 5 ( ‘1 xn= l/2. 2n Thus, with probability at least l/2, no ele- ment is singular. The lemma follows from 3. The Isolating Lemma the observation that there is a unique minimum weight set iff no element is singu- lar. Definition: A set system (S, F) consists of a Notice that by the same argument, the finite set S of elements, maximum weight set will be unique with pro- s = {Xl, x2, . ..) x,), and a family F of sub- bability at least l/2 as well. An extension of sets of S, i.e. F = {S,, S2, * - - Sk}, SQ, the isolating lemma is required in Theorem for 1 5 j I k. 2(a). In this extension we are given integers Let us assign a weight wi to each ele- al,a2r ’ ’ ’ at, and the weight of set Si is ment Xi E S and let us define the weight of defined to be aj + C Wi, 15 j5 k. The the set Sj to be C Wi. x&sj proof given above works for this case as well. Lemma 1: Let (S, F) be a set system whose elements are assigned integer weights chosen uniformly and independently from 11, 2n], Then, Pr [There is a unique minimum weight set 4. The Matching Algorithm 1 in Fl2 2’ We will first consider the simpler case of a bipartite graph: Proof: Fix the weights of all elements except Xi. Define the threshold for element Xi, to be Input: A bipartite graph G(iJ,V,E), having the real number a; such that if wi I oi then a perfect matching. Xi is contained in some minimum weight sub- set, Si, and if wi > oi then Xi is in no Problem : Find a perfect matching in G. minimum weight subset. We will view the edges in E and the Clearly, if Wi < oi, then the element Xi set of perfect matchings in G as a set system. must be in every minimum weight subset. Let us assign random integer weights to the Thus ambiguity about element Xi occurs iff edges of the graph, chosen uniformly and Wi = (xi, since in this case there is a independently from [I, 2m1, where minimum weight subset that contains Xi and m = [El. Now by lemma 1, the minimum another which does not. In this case we shall weight perfect matching in G will be unique say that the element Xi is singular. with probability at least l/2. Our parallel We now make the crucial observation algorithm will pick out this perfect matching.
Load more

Recommended publications
  • Planar Graph Perfect Matching Is in NC
    2018 IEEE 59th Annual Symposium on Foundations of Computer Science Planar Graph Perfect Matching is in NC Nima Anari Vijay V. Vazirani Computer Science Department Computer Science Department Stanford University University of California, Irvine [email protected] [email protected] Abstract—Is perfect matching in NC? That is, is there a finding a perfect matching, was obtained by Karp, Upfal, and deterministic fast parallel algorithm for it? This has been an Wigderson [11]. This was followed by a somewhat simpler outstanding open question in theoretical computer science for algorithm due to Mulmuley, Vazirani, and Vazirani [17]. over three decades, ever since the discovery of RNC matching algorithms. Within this question, the case of planar graphs has The matching problem occupies an especially distinguished remained an enigma: On the one hand, counting the number position in the theory of algorithms: Some of the most of perfect matchings is far harder than finding one (the former central notions and powerful tools within this theory were is #P-complete and the latter is in P), and on the other, for discovered in the context of an algorithmic study of this NC planar graphs, counting has long been known to be in problem, including the notion of polynomial time solvability whereas finding one has resisted a solution. P In this paper, we give an NC algorithm for finding a perfect [4] and the counting class # [22]. The parallel perspective matching in a planar graph. Our algorithm uses the above- has also led to such gains: The first RNC matching algorithm stated fact about counting matchings in a crucial way.
    [Show full text]
  • Structural Results on Matching Estimation with Applications to Streaming∗
    Structural Results on Matching Estimation with Applications to Streaming∗ Marc Bury, Elena Grigorescu, Andrew McGregor, Morteza Monemizadeh, Chris Schwiegelshohn, Sofya Vorotnikova, Samson Zhou We study the problem of estimating the size of a matching when the graph is revealed in a streaming fashion. Our results are multifold: 1. We give a tight structural result relating the size of a maximum matching to the arboricity α of a graph, which has been one of the most studied graph parameters for matching algorithms in data streams. One of the implications is an algorithm that estimates the matching size up to a factor of (α + 2)(1 + ") using O~(αn2=3) space in insertion-only graph streams and O~(αn4=5) space in dynamic streams, where n is the number of nodes in the graph. We also show that in the vertex arrival insertion-only model, an (α + 2) approximation can be achieved using only O(log n) space. 2. We further show that the weight of a maximum weighted matching can be ef- ficiently estimated by augmenting any routine for estimating the size of an un- weighted matching. Namely, given an algorithm for computing a λ-approximation in the unweighted case, we obtain a 2(1+")·λ approximation for the weighted case, while only incurring a multiplicative logarithmic factor in the space bounds. The algorithm is implementable in any streaming model, including dynamic streams. 3. We also investigate algebraic aspects of computing matchings in data streams, by proposing new algorithms and lower bounds based on analyzing the rank of the Tutte-matrix of the graph.
    [Show full text]
  • The Polynomially Bounded Perfect Matching Problem Is in NC2⋆
    The polynomially bounded perfect matching problem is in NC2⋆ Manindra Agrawal1, Thanh Minh Hoang2, and Thomas Thierauf3 1 IIT Kanpur, India 2 Ulm University, Germany 3 Aalen University, Germany Abstract. The perfect matching problem is known to be in ¶, in ran- domized NC, and it is hard for NL. Whether the perfect matching prob- lem is in NC is one of the most prominent open questions in complexity theory regarding parallel computations. Grigoriev and Karpinski [GK87] studied the perfect matching problem for bipartite graphs with polynomially bounded permanent. They showed that for such bipartite graphs the problem of deciding the existence of a perfect matchings is in NC2, and counting and enumerating all perfect matchings is in NC3. For general graphs with a polynomially bounded number of perfect matchings, they show both problems to be in NC3. In this paper we extend and improve these results. We show that for any graph that has a polynomially bounded number of perfect matchings, we can construct all perfect matchings in NC2. We extend the result to weighted graphs. 1 Introduction Whether there is an NC-algorithm for testing if a given graph contains a perfect matching is an outstanding open question in complexity theory. The problem of deciding the existence of a perfect matching in a graph is known to be in ¶ [Edm65], in randomized NC2 [MVV87], and in nonuniform SPL [ARZ99]. This problem is very fundamental for other computational problems (see e.g. [KR98]). Another reason why a derandomization of the perfect matching problem would be very interesting is, that it is a special case of the polynomial identity testing problem.
    [Show full text]
  • 3.1 the Existence of Perfect Matchings in Bipartite Graphs
    15-859(M): Randomized Algorithms Lecturer: Shuchi Chawla Topic: Finding Perfect Matchings Date: 20 Sep, 2004 Scribe: Viswanath Nagarajan 3.1 The existence of perfect matchings in bipartite graphs We will look at an efficient algorithm that determines whether a perfect matching exists in a given bipartite graph or not. This algorithm and its extension to finding perfect matchings is due to Mulmuley, Vazirani and Vazirani (1987). The algorithm is based on polynomial identity testing (see the last lecture for details on this). A bipartite graph G = (U; V; E) is specified by two disjoint sets U and V of vertices, and a set E of edges between them. A perfect matching is a subset of the edge set E such that every vertex has exactly one edge incident on it. Since we are interested in perfect matchings in the graph G, we shall assume that jUj = jV j = n. Let U = fu1; u2; · · · ; ung and V = fv1; v2; · · · ; vng. The algorithm we study today has no error if G does not have a perfect matching (no instance), and 1 errs with probability at most 2 if G does have a perfect matching (yes instance). This is unlike the algorithms we saw in the previous lecture, which erred on no instances. Definition 3.1.1 The Tutte matrix of bipartite graph G = (U; V; E) is an n × n matrix M with the entry at row i and column j, 0 if(ui; uj) 2= E Mi;j = xi;j if(ui; uj) 2 E The determinant of the Tutte matrix is useful in testing whether a graph has a perfect matching or not, as the following claim shows.
    [Show full text]
  • Chapter 4 Introduction to Spectral Graph Theory
    Chapter 4 Introduction to Spectral Graph Theory Spectral graph theory is the study of a graph through the properties of the eigenvalues and eigenvectors of its associated Laplacian matrix. In the following, we use G = (V; E) to represent an undirected n-vertex graph with no self-loops, and write V = f1; : : : ; ng, with the degree of vertex i denoted di. For undirected graphs our convention will be that if there P is an edge then both (i; j) 2 E and (j; i) 2 E. Thus (i;j)2E 1 = 2jEj. If we wish to sum P over edges only once, we will write fi; jg 2 E for the unordered pair. Thus fi;jg2E 1 = jEj. 4.1 Matrices associated to a graph Given an undirected graph G, the most natural matrix associated to it is its adjacency matrix: Definition 4.1 (Adjacency matrix). The adjacency matrix A 2 f0; 1gn×n is defined as ( 1 if fi; jg 2 E; Aij = 0 otherwise. Note that A is always a symmetric matrix with exactly di ones in the i-th row and the i-th column. While A is a natural representation of G when we think of a matrix as a table of numbers used to store information, it is less natural if we think of a matrix as an operator, a linear transformation which acts on vectors. The most natural operator associated with a graph is the diffusion operator, which spreads a quantity supported on any vertex equally onto its neighbors. To introduce the diffusion operator, first consider the degree matrix: Definition 4.2 (Degree matrix).
    [Show full text]
  • Algebraic Graph Theory
    Algebraic graph theory Gabriel Coutinho University of Waterloo November 6th, 2013 We can associate many matrices to a graph X . Adjacency: 0 0 1 0 1 0 1 B 1 0 1 0 1 C B C A(X ) = B 0 1 0 1 1 C B C @ 1 0 1 0 0 A 0 1 1 0 0 Tutte: 0 1 1 0 0 0 0 1 0 1 0 x12 0 x14 0 1 0 1 1 0 0 B C B −x12 0 x23 0 x25 C Incidence: B 0 0 1 0 1 1 C B C B C B 0 −x23 0 x34 x35 C B 0 1 0 0 0 1 C B C @ A @ −x14 0 −x34 0 0 A 0 0 0 1 1 0 0 −x25 −x35 0 0 What is it about? - the tools Matrices Problem 1 - (very) regular graphs Groups Problem 2 - quantum walks Polynomials Matrices Gabriel Coutinho Algebraic graph theory 2 / 30 Adjacency: 0 0 1 0 1 0 1 B 1 0 1 0 1 C B C A(X ) = B 0 1 0 1 1 C B C @ 1 0 1 0 0 A 0 1 1 0 0 Tutte: 0 1 1 0 0 0 0 1 0 1 0 x12 0 x14 0 1 0 1 1 0 0 B C B −x12 0 x23 0 x25 C Incidence: B 0 0 1 0 1 1 C B C B C B 0 −x23 0 x34 x35 C B 0 1 0 0 0 1 C B C @ A @ −x14 0 −x34 0 0 A 0 0 0 1 1 0 0 −x25 −x35 0 0 What is it about? - the tools Matrices Problem 1 - (very) regular graphs Groups Problem 2 - quantum walks Polynomials Matrices We can associate many matrices to a graph X .
    [Show full text]
  • Perfect Matching Testing Via Matrix Determinant 1 Polynomial Identity 2
    MS&E 319: Matching Theory Instructor: Professor Amin Saberi ([email protected]) Lecture 1: Perfect Matching Testing via Matrix Determinant 1 Polynomial Identity Suppose we are given two polynomials and want to determine whether or not they are identical. For example, (x 1)(x + 1) =? x2 1. − − One way would be to expand each polynomial and compare coefficients. A simpler method is to “plug in” a few numbers and see if both sides are equal. If they are not, we now have a certificate of their inequality, otherwise with good confidence we can say they are equal. This idea is the basis of a randomized algorithm. We can formulate the polynomial equality verification of two given polynomials F and G as: H(x) F (x) G(x) =? 0. ≜ − Denote the maximum degree of F and G as n. Assuming that F (x) = G(x), H(x) will be a polynomial of ∕ degree at most n and therefore it can have at most n roots. Choose an integer x uniformly at random from 1, 2, . , nm . H(x) will be zero with probability at most 1/m, i.e the probability of failing to detect that { } F and G differ is “small”. After just k repetitions, we will be able to determine with probability 1 1/mk − whether the two polynomials are identical i.e. the probability of success is arbitrarily close to 1. The following result, known as the Schwartz-Zippel lemma, generalizes the above observation to polynomials on more than one variables: Lemma 1 Suppose that F is a polynomial in variables (x1, x2, .
    [Show full text]
  • Planar Graph Perfect Matching Is in NC
    Planar Graph Perfect Matching is in NC Nima Anari1 and Vijay V. Vazirani2 1Computer Science Department, Stanford University,∗ [email protected] 2Computer Science Department, University of California, Irvine,y [email protected] Abstract Is perfect matching in NC? That is, is there a deterministic fast parallel algorithm for it? This has been an outstanding open question in theoretical computer science for over three decades, ever since the discovery of RNC matching algorithms. Within this question, the case of planar graphs has remained an enigma: On the one hand, counting the number of perfect matchings is far harder than finding one (the former is #P-complete and the latter is in P), and on the other, for planar graphs, counting has long been known to be in NC whereas finding one has resisted a solution. In this paper, we give an NC algorithm for finding a perfect matching in a planar graph. Our algorithm uses the above-stated fact about counting matchings in a crucial way. Our main new idea is an NC algorithm for finding a face of the perfect matching polytope at which W(n) new conditions, involving constraints of the polytope, are simultaneously satisfied. Several other ideas are also needed, such as finding a point in the interior of the minimum weight face of this polytope and finding a balanced tight odd set in NC. 1 Introduction Is perfect matching in NC? That is, is there a deterministic parallel algorithm that computes a perfect matching in a graph in polylogarithmic time using polynomially many processors? This has been an outstanding open question in theoretical computer science for over three decades, ever since the discovery of RNC matching algorithms [KUW86; MVV87].
    [Show full text]
  • 1 Matchings in Graphs
    CS367 Lecture 8 Perfect Matching Scribe: Amir Abboud Date: April 28, 2014 1 Matchings in graphs This week we will be talking about finding matchings in graphs: a set of edges that do not share endpoints. Definition 1.1 (Maximum Matching). Given an undirected graph G =(V,E), find a subset of edges M ⊆ E of maximum size such that every pair of edges e,e′ ∈ M do not share endpoints e ∩ e′ = ∅. Definition 1.2 (Perfect Matching). Given an undirected graph G = (V,E) where |V | = n is even, find a subset of edges M ⊆ E of size n/2 such that every pair of edges e,e′ ∈ M do not share endpoints e ∩ e′ = ∅. That is, every node must be covered by the matching M. Obviously, any algorithm for Maximum Matching gives an algorithm for Perfect Matching. It is an exercise to show that if one can solve Perfect Matching in T (n) time, then one can solve Maximum Matching in time O˜(T (2n)). The idea is to binary search for the maximum k for which there is a matching M with |M|≥ k. To check whether such M exists, we can add a clique on n − 2k nodes to the graph and connect it to the original graph with all possible edges. The new graph will have a perfect matching if and only if the original graph had a matching with k edges. We will focus on Perfect Matching and give algebraic algorithms for it. Because of the above reduction, this will also imply algorithms for Maximum Matching.
    [Show full text]
  • Matrix Multiplication Based Graph Algorithms
    Matrix Multiplication and Graph Algorithms Uri Zwick Tel Aviv University February 2015 Last updated: June 10, 2015 Short introduction to Fast matrix multiplication Algebraic Matrix Multiplication j i Aa ()ij Bb ()ij = Cc ()ij Can be computed naively in O(n3) time. Matrix multiplication algorithms Complexity Authors n3 — n2.81 Strassen (1969) … n2.38 Coppersmith-Winograd (1990) Conjecture/Open problem: n2+o(1) ??? Matrix multiplication algorithms - Recent developments Complexity Authors n2.376 Coppersmith-Winograd (1990) n2.374 Stothers (2010) n2.3729 Williams (2011) n2.37287 Le Gall (2014) Conjecture/Open problem: n2+o(1) ??? Multiplying 22 matrices 8 multiplications 4 additions Works over any ring! Multiplying nn matrices 8 multiplications 4 additions T(n) = 8 T(n/2) + O(n2) lg8 3 T(n) = O(n )=O(n ) ( lgn = log2n ) “Master method” for recurrences 푛 푇 푛 = 푎 푇 + 푓 푛 , 푎 ≥ 1 , 푏 > 1 푏 푓 푛 = O(푛log푏 푎−휀) 푇 푛 = Θ(푛log푏 푎) 푓 푛 = O(푛log푏 푎) 푇 푛 = Θ(푛log푏 푎 log 푛) 푓 푛 = O(푛log푏 푎+휀) 푛 푇 푛 = Θ(푓(푛)) 푎푓 ≤ 푐푛 , 푐 < 1 푏 [CLRS 3rd Ed., p. 94] Strassen’s 22 algorithm CABAB 11 11 11 12 21 MAABB1( 11 Subtraction! 22 )( 11 22 ) CABAB12 11 12 12 22 MAAB2() 21 22 11 CABAB21 21 11 22 21 MABB3 11() 12 22 CABAB22 21 12 22 22 MABB4 22() 21 11 MAAB5() 11 12 22 CMMMM 11 1 4 5 7 MAABB6 ( 21 11 )( 11 12 ) CMM 12 3 5 MAABB7( 12 22 )( 21 22 ) CMM 21 2 4 7 multiplications CMMMM 22 1 2 3 6 18 additions/subtractions Works over any ring! (Does not assume that multiplication is commutative) Strassen’s nn algorithm View each nn matrix as a 22 matrix whose elements
    [Show full text]
  • Matchings in Graphs 1 Definitions 2 Existence of a Perfect Matching
    Matchings in Graphs Lecturer: Prajakta Nimbhorkar Meeting: 7 Scribe: Karteek 18/02/2010 In this lecture we look at an \efficient" randomized parallel algorithm for each of the follow- ing: • Checking the existence of perfect matching in a bipartite graph. • Checking the existence of perfect matching in a general graph. • Constructing a perfect matching for a given graph. • Finding a min-weight perfect matching in the given weighted graph. • Finding a Maximum matching in the given graph. where “efficient" means an NC-algorithm or an RNC-algorithm. 1 Definitions Definition 1 NC: The set of problems that can be solved within poly-log time by a polynomial number of processors working in parallel. Definition 2 RNC: The set of problems for which there is a poly-log time algorithm which uses polynomially many processors in parallel and in addition is allowed polynomially many random bits and the following properties hold: 1 • If the correct answer is YES, the algorithm returns YES with probabilty ≥ 2 . • If the correct answer is NO, the algorithm always returns NO. 2 Existence of a perfect matching 2.1 Bipartite Graphs Here we consider the decision problem of checking if a bipartite graph has a perfect matching. Let the given graph be G = (V; E). Let V =A[B. Since we are looking for perfect matchings, we only look at graphs where jAj = jBj = n, because otherwise, there cannot be a perfect matching. Fact 3 In a bipartite graph as above, every perfect matching can be seen as a permutation from Sn and every permutation from Sn can be seen as perfect matching.
    [Show full text]
  • 1 Introduction 2 Improving Rabin-Vazirani To
    Lecture 10 More on Matching Scribe: Michael P. Kim Date: April 30, 2014 1 Introduction In the last lecture, we discussed two algorithms for finding perfect matchings both of which leveraged a random substitution of the Tutte Matrix to find edges in a pefect matching. Recall the Tutte Matrix is defined as follows. 0 if(i,j) 6∈ E T (i,j)= xij if i<j −xij if i>j The naive algorithm uses the determinant of the Tutte Matrix as an oracle for whether a perfect matching exists, while the Rabin-Vazirani Algorithm finds edges to include in the matching by leveraging properties of the Tutte Matrix. Recall the naive algorithm. Algorithm 1: NaiveMatch(G) for e ∈ E do if det TG\{e} 6= 0 then Remove e from G; Because we can compute the determinant of T with an O˜(nω) randomized algorithm, this gives us an O(nω+2) benchmark. Now, recall the Rabin -Vazirani algorithm. Algorithm 2: RV(G) T ← random substitution of Tutte Matrix mod some prime p > n3; M ← ∅; n while |M| < 2 do N ← T −1 // run time bottleneck to invert T Find j s.t. N(1,j) 6= 0, T (1,j) 6= 0; M ← M ∪ {(i,j)}; T ← T{1,j},{1,j} Remember that the runtime of RV is O(nω+1), primarily due to the bottleneck in the while loop to compute the inverse of the Tutte Matrix in each iteration. In this lecture, we will improve upon this run time, to O(n3) by using a specialized algorithm to invert the Tutte Matrix.
    [Show full text]