On Dominating Sets for Pseudo-disks
|
|
- Jonathan Jacobs
- 6 years ago
- Views:
Transcription
1 On Dominating Sets for Pseudo-disks Boris Aronov 1, Anirudh Donakonda 1, Esther Ezra 2,, and Rom Pinchasi 3 1 Department of Computer Science and Engineering Tandon School of Engineering, New York University Brooklyn, NY 11201, USA boris.aronov@nyu.edu, ad2930@nyu.edu. 2 Department of Mathematics Georgia Institute of Technology Atlanta, Georgia 30332, USA eezra3@math.gatech.edu. 3 Department of Mathematics, Technion - Israel Institute of Technology, Haifa, Israel room@math.technion.ac.il. Abstract. We study the problem of computing the smallest weighted dominating set in a collection of pseudo-disks in the plane, i.e., a minimumsize subset of the given set of pseudo-disks so that every pseudo-disk is either selected or intersected by a selected pseudo-disk. A similar problem has recently been investigated by Gibson and Pirwani (ESA 2010) for disk graphs, whereas the case of pseudo-disks remains open. Using the machinery of Chan et al. (SODA 2012) we present a randomized expected polynomial-time algorithm that achieves a constant approximation factor of the smallest weighted dominating set in a collection of pseudo-disks. 1 Introduction Problem statement. We are given a finite collection P of pseudo-disks in the plane, i.e., a collection of compact regions each bounded by a Jordan curve such that any two of their boundaries cross twice or not at all. We define the intersection graph G of P in the standard manner, that is, the vertex set is P and there is an edge between two pseudo-disks if their intersection is non-empty. Work on this paper by Boris Aronov has been supported by NSA MSP Grant H , by NSF Grants CCF , CCF , CCF , and CCF , and by BSF grant 2014/170. Work on this paper by Anirudh Donakonda has been partially supported by NSF Grant CCF Work on this paper by Esther Ezra has been supported by NSF under grants CAREER CCF , CCF , and CCF Work on this paper by Esther Ezra was initiated when she was at the Courant Institute of Mathematical Sciences, New York University, New York, NY 10012, USA, and continued at the Department of Computer Science and Engineering, Polytechnic (now Tandon) School of Engineering, New York University, Brooklyn, NY 11201, USA.
2 The dominating set problem for G is to find a smallest subset D P, such that each vertex in G is either in D or is adjacent to a vertex in D. In other words, this is a smallest subset of P such that any pseudo-disk in P appears in the subset or is intersected by a pseudo-disk in it. In the weighted dominating set problem, each element of P is assigned a non-negative weight, and the goal is to find a dominating set of overall smallest weight. Related work. For general graphs, solving the dominating set problem is NP-hard [14, 20], and the standard greedy algorithm yields an O(1 + log n)- approximation factor [7, 19], where n is the size of the vertex set, which, up to a constant factor, is the best one can do in polynomial-time time, under reasonable complexity-theoretic assumptions [6, 10]. Many graph problems remain NP-hard in more specialized settings, such as unit disk graphs and growth-bounded graphs [8]. Nevertheless, often one can achieve in polynomial time better approximation factors. Specifically, the dominating set problem admits a polynomial-time approximation scheme (PTAS) for the aforementioned settings [18, 23]; see also [12] for a constant-factor approximation for the weighted dominating set problem on unit disk graphs. Erlebach and van Leeuwen [11] studied the dominating set problem in several natural cases, where they constructed polynomial-time algorithms guaranteeing a constant approximation factor. This includes (i) regular r-gons, where r is an arbitrary constant, (ii) homothetic triangles, and (iii) axis-parallel rectangles of bounded aspect ratio. On the other hand, they showed that in intersection graphs of arbitrary fat convex objects (a fat object is one in which the ratio of the radius of the smallest circumscribing disk to that of the largest inscribed disk is bounded by a constant), of arbitrarily large complexity, the problem is as hard to approximate as in abstract settings, and when the algebraic description complexity of these objects is constant, 4 then the problem is APX-hard. 5 In addition, they showed that for axis-parallel rectangles of arbitrary aspect ratio the problem is APX-hard as well. This result settles a conjecture of Chlebík and Chlebíková [6], who showed APX-hardness for axis-parallel boxes in any dimension d 3. See also the recent paper of Har-Peled and Quanrud [17] studying intractability in geometric settings. Gibson and Pirwani studied the case of disk graphs, that is, the special case of our problem when one replaces P with a set of Euclidean disks [15]. In this situation they demonstrated a PTAS for computing the smallest (unweighted) dominating set via a local search approach. For the weighted problem, they showed that 2 O(log n) -approximation factor is achievable in polynomial time using the quasi-sampling approach of Varadarajan [26]. The latter has been strengthened to O(1) by Chan et al. [5], who improved the sampling scheme of Varadarajan [26] and also presented a simple reduction from set cover to dominating set, considerably simplifying the approach taken in [15]. 4 A planar region is of constant (algebraic description) complexity if it can be described by a Boolean formula constructed from at most c algebraic inequalities in x and y, of total degree at most c, for some constant c. 5 This excludes the case of disks discussed below.
3 Our result. We show that, for the case of pseudo-disks, each of which is assigned a non-negative weight, one can achieve a constant approximation factor in expected polynomial time. In our analysis we define an appropriate set system for P, which, roughly speaking, consists of subsets of P which are met by a common pseudodisk (see Section 2 for the precise definition). Our main technical contribution is to show that this set system has VC-dimension of at most 4 (Theorem 3), and, moreover, that it has shallow cell complexity, meaning that the number of sets of size k P is only linear in P and polynomial in k (Definition 1 and Proposition 1). The aforementioned approximation factor is then achieved by using the recent machinery of Chan et al. [5], which allows us to deduce the following main result. Theorem 1. There is a randomized expected polynomial-time algorithm, that given a set P of pseudo-disks in the plane, each with a non-negative weight, computes a dominating set D P of weight O(Opt), where Opt is the smallest total weight of such dominating set. We note that the analysis of Gibson and Pirwani [15], as well as the more recent improvement of Chan et al. [5] strongly rely on the geometry of the disks. This enables them to reduce the dominating set problem for disks to a hitting set or a set cover for other geometric objects. However, neither of the approaches in [5, 15] seems to extend to the case of pseudo-disks, which is the reason we had to resort to another set of tools that largely exploit the combinatorial structure of the setting, in contrast to the more geometric ones used in [5, 15]. 2 Preliminaries Arrangements and levels. Let P be a set of n pseudo-disks in the plane. Without loss of generality, we assume that the pseudo-disks of P are in general position, that is, no point is incident to more than two pseudo-disk boundaries and, whenever two such boundaries meet, they properly cross. Let A(P) denote the arrangement of P (see, e.g., [1]). The level of an (open) face in this arrangement is the number of pseudo-disks containing it in their interior. Well-known results by Kedem et al. [21] and by Clarkson and Shor [9] imply that A(P) has O(n) level-2 faces and O(nk) faces at level at most k. Range spaces of bounded VC-dimension. A range space (or set system) (X, R) is a pair consisting of an underlying universe X of objects (also called the space) and a family R of subsets of X (called ranges). Of particular interest are range spaces of bounded VC-dimension, as, e.g., their restriction to any finite subset X X admits only polynomially many (in X ) ranges (see, e.g., [16, Chapter 5]). They also have other structural properties, which we discuss shortly. We say a subset K X is shattered by R if, for every subset Z of K, Z = K r for some range r R. The VC-dimension of a range space is the size of the largest finite shattered subset; if no such subset exists one says the VC-dimension
4 P 2 P 3 is infinite, slightly abusing the terminology. If the dimension is bounded by an absolute constant d > 0, we say the range space has bounded VC-dimension, again somewhat abusing the terminology. In this case, the number of ranges in any restriction of (X, R) to a finite collection X X is at most O( X d ) (once again, see [16, Chapter 5]). Hitting sets and dominating sets. Fix any family F of pseudo-disks in the plane. We first define the appropriate range spaces, which correspond to our dominating set problem. Namely, we consider finite range spaces (P, E), where P F is a finite set of pseudo-disks in the plane, and each range r E is precisely the set of pseudo-disks of P intersected by a pseudo-disk of F (which may not belong to P). Consider the intersection graph G of P as defined above. In G, a neighborhood of a pseudo-disk is the set of pseudo-disks intersecting it (therefore, this is a subgraph of G spanned by a star; note that we include the pseudo-disk itself in its neighborhood). The family of all neighborhoods define a range space, arising as a special case of the range space (P, E) defined above, as in this case we have F = P. In what follows and with a slight abuse of notation, we also denote a range space of the latter kind by (P, E). In fact, for our purposes it is sufficient to consider only the case F = P. However, we do not make this assumption in our analysis in Sections 3 and 4. We now observe that a dominating set in G is, in fact, a hitting set for (P, E), where the latter refers to a subset D P, which meets all ranges of E. That is, D meets all objects in P if and only if each neighborhood in the intersection graph (which is a range of E) is hit by an element of D. In particular, the minimum hitting set for (P, E) corresponds to the minimum dominating set of G, and this property holds in the weighted setting as well. See Figure 1 for an example. P 2 P 3 P 6 P 1 P 4 P 1 P 5 (a) P 6 P 5 (b) P 4 Fig. 1. The intersection graph in (b) induced by the pseudo-disks P 1,..., P 6, depicted in (a), is a star centered at P 1. The smallest dominating set is {P 1}. The underlying range space is (P, E), where P = {P 1,..., P 6}, and E = {{P 1,..., P 6}, {P 1, P 2}, {P 1, P 3}, {P 1, P 4}, {P 1, P 5}, {P 1, P 6}}. Finally, {P 1} is the smallest hitting set for this range space.
5 Shallow cell complexity. Following the notation of Chan et al. [5], we represent our set system (P, E) as an incidence matrix A. The columns of A correspond to pseudo-disks of P, and the rows to the sets in E. We have A i,j = 1 if the set e i E contains the pseudo-disk p j P, and 0 otherwise, for i = 1,..., E, j = 1,..., P. In a similar manner we can represent any restriction of (P, E) to a subset P P by considering a submatrix A of A restricted to the corresponding columns. In this manner, we may obtain sets of equivalent rows containing the same 0/1 vector. In the terminology of Chan et al. [5], equivalent classes of rows are referred to as cells, and the depth of a cell is the total number of ones contained in any one of its rows. We next define the notion of shallow cell complexity : Definition 1 (Shallow Cell Complexity; Chan et al. [5]). Let f(n, k) be a non-decreasing function in both n and k. A binary matrix A with N columns has shallow cell complexity (SCC) f if for all 1 k n N and for all submatrices A of A containing exactly n columns, the number of cells of A of depth k or fewer is at most f(n, k). A hitting set instance has SCC f if and only if its element-set incidence matrix does. 6 A main result of Chan et al. [5], which we employ in this study, is that hittingset instances with small SCC admit small approximation factors achievable in expected polynomial time: Theorem 2 (Chan et al. [5]). If I is a class of hitting set instances with SCC f(n, k) = c 1 nk c2, for absolute constants c 1, c 2 0, then there exists a randomized polynomial-time O(1)-approximation algorithm for the weighted hitting set problem for I. We note that the original formulation of Theorem 2 is slightly more general, and covers the possibility that the dependence of f on n is slightly super-linear. However, for our purposes a linear dependence on n is sufficient. In the main result of this study (Proposition 1) we show that the set system (P, E) admits SCC with f(n, k) = O(nk 3 ). Then, Theorem 1 follows by applying Theorem 2, and using our observation that the minimum weight hitting set for (P, E) corresponds to the minimum weight dominating set of G. Note that we need to assume that the intersection predicate for two pseudo-disks can be evaluated in time polynomial in the sum of their sizes, which enables us to produce the incidence matrix A of (P, E) in polynomial time, and therefore eventually apply the algorithm in [5] in overall expected polynomial time. Thus our main goal is to show that (P, E) has small shallow cell complexity. Road map. The rest of the paper is organized as follows. In Section 3 we prove that VC-dimension of (P, E) is at most 4 and that this bound is optimal (Theorem 3). In Section 4 we show that the number of pairs in E (sets containing exactly 6 The original definitions and analysis in [5] are stated using set cover terminology. We will use the dual and equivalent hitting set formulation, as it is more appropriate for our application.
6 two elements of P) is only linear in P, which is a consequence of the analysis concerning the VC-dimension. Then we show in Section 4 that the number of triples and quadruples in E is linear in P as well. Our proof technique exploits several ideas from the work of Buzaglo et al. [4], who studied the corresponding hitting-set problem for points and pseudo-disks in the plane. Our analysis is a variant of that of Buzaglo et al. [4], but our setting is more general. Specifically, Buzaglo et al. [4] studied topological hypergraphs defined by points and pseudo-disks enclosing them, whereas we consider a hypergraph of pseudo-disks, and subsets of them intersected by pseudo-disks. By viewing a point to be a sufficiently small pseudo-disk one observes that our result is a non-trivial generalization of that in [4]. Finally, we comment that it is crucial to consider pseudo-disks rather than pseudo-circles (that is, full regions rather than their boundaries). Indeed, range spaces of pseudo-circles and subsets of them met by a pseudo-circle do not satisfy Theorem 4: Consider n pairwise intersecting circles in general position and for each of the ( n 2) pairs of circles, place a tiny circle at one of their intersection points. This yields a collection of quadratically many circle pairs, rather than a linear number asserted in Theorem 4. 3 The VC-Dimension is Bounded In this section we prove: Theorem 3. A range space (P, E) as defined above has VC-dimension at most 4. This bound is the best possible. We start by stating the following technical lemma from [4] (see figure below): γ D γ D Lemma 1 (Buzaglo et al. [4]). Let γ and γ be arbitrary curves contained in pseudo-disks D and D, respectively. If the endpoints of γ lie outside of D and the endpoints of γ lie outside of D, then γ and γ must cross an even number of times. We say that a set K of pseudo-disks is well behaved if every pseudo-disk in K has a point not covered by the union of other pseudo-disks in K. We begin with an auxiliary construction. Let K be a finite well-behaved set of pseudo-disks. We construct a graph H = H(K) whose vertices correspond to pseudo-disks of K and whose edges correspond to pseudo-disks of F that meet precisely two sets of K. More specifically, we draw H as follows:
7 Vertices of H: For each pseudo-disk D K, we fix a point v(d) (which need not lie on the boundary of D), not contained in any other pseudo-disk of K; it exists since K is well behaved. The points {v(d) : D K} form the vertex set of H. Edges of H: Let D 1, D 2 K, v 1 = v(d 1 ) and v 2 = v(d 2 ). Suppose there exists P F that intersects D 1 and D 2 but not any other disk in K; fix one such P. Notice that it is possible that P K. We will add an edge v 1 v 2 to H, drawn as described below. We call a connected portion of the edge inside P a red arc and such a portion outside P a blue arc. The edge v 1 v 2 consists of at most one red arc and at most two blue arcs. In figures below, we also use the convention of drawing pseudo-disks of K in blue and the connecting pseudo-disk(s) in red. P contains both v 1 and v 2 : Draw a red arc in P from v 1 to v 2. This forms the edge v 1 v 2 H. See figure below. v 2 D 1 γ D 2 v 1 P P v 2 D 1 v D 2 1 γ P contains v 1 but does not contain v 2 : Draw a red arc in P that starts at v 1 and ends at the boundary of P inside D 2. Now draw a blue arc in D 2 that starts at this point, ends at v 2 and lies totally outside P. The concatenation of these two arcs forms the edge v 1 v 2 of H. See figure below. P D 1 D 2 D 2 D 1 v γ γ 1 v v 2 2 v 1 P D 1 =P v 1 D 2 γ v 2 P contains neither v 1 nor v 2 : Draw a blue arc in D 1 that starts at v 1, ends at the boundary of P inside D 1, and otherwise stays outside of P. From its endpoint, draw a red arc in P to a point of the boundary of P inside D 2 and from there, draw the final blue arc outside P in D 2 to the vertex v 2. The concatenation of these three arcs constitutes the edge v 1 v 2. See figure below. D 1 D 2 P γ v 2 v 1 v 1 D 1 P γ D 2 v 2
8 By construction, for each arc of the constructed edge, either red or blue, there is a pseudo-disk that completely contains it. Lemma 2. The graph H = H(K) is planar. Proof. We will prove H is planar using the strong Hanani-Tutte theorem [25]. Consider two edges e, e that connect v 1 = v(d 1 ) to v 2 = v(d 2 ), and v 3 = v(d 3 ) to v 4 = v(d 4 ) in H(K), respectively, and do not share a vertex so that D 1, D 2, D 3, D 4 K are pairwise distinct. We will prove that e and e intersect an even number of times, by considering their red and blue portions separately. Let P 12 F be the pseudo-disk intersecting only D 1 and D 2 that was used to draw e and let P 34 F be the corresponding pseudo-disk intersecting only D 3 and D 4. Red-Blue Intersections: Consider the red portion of e. This red arc is contained in P 12 and therefore does not meet any pseudo-disk of K other than D 1, D 2. As the blue portions of e lie inside D 3, D 4, this implies that the red arc of e does not meet the blue portions of e. Symmetrically the red portion of e cannot intersect the blue portions of e. Red-Red Intersections: The red arc α along e lies entirely in P 12 and has one endpoint in D 1 and the other in D 2. Similarly, the red arc α along e lies entirely in P 34 and has one endpoint in D 3 and the other in D 4. As P 12 does not intersect D 3 and D 4 and P 34 does not intersect D 1 and D 2, the endpoints of α do not lie in P 34 and the endpoints of α do not lie in P 12. By Lemma 1, α and α intersect an even number of times. Blue-Blue Intersections: Consider blue arcs β e and β e. The blue arc β starts, say, at vertex v 1 of pseudo-disk D 1 and ends at p in D 1 on the boundary of pseudo-disk P 12, and β starts, say, at vertex v 3 of pseudo-disk D 3 and ends at p in D 3 on the boundary of pseudo-disk P 34. By the construction of the vertices of H, v 1 / D 3 and v 3 / D 1. Now, p cannot lie in D 3 because P 12 meets only D 1 and D 2 and similarly p cannot lie in D 1. Hence, by Lemma 1 we deduce once again that β and β intersect an even number of times. There is a possibility that some edges of H self-intersect, but such intersections can be removed using standard methods: see for example [24], and Figure 2 in Appendix A. Thus, any two edges of H that do not share an endpoint cross an even number of times, and therefore H is planar by the strong Hanani-Tutte theorem [25]. Proof (of Theorem 3). Let K P be a set shattered by E. Since K is shattered, for every disk P K there is a disk F F that intersects P but not other disk in K. Therefore, K is well behaved. For a well-behaved set K, H(K) is planar, by Lemma 2, and therefore has at most 3 K 6 edges. However, K is shattered, so H(K) is a complete graph and has ( ) ( K 2 edges. Therefore, K ) 2 3 K 6, implying K 4. This means that the VC-dimension of (P, E) is at most 4. Refer to Figure 3 in Appendix B that shows that this bound is tight.
9 4 Bounding the Number of Shallow Sets Given a family F of pseudo-disks in the plane and a subfamily P F, recall that we denote by (P, E) the hypergraph (or set system) whose vertices are the pseudo-disks in P and whose edges (or sets) are the subsets of pseudo-disks in P intersected by a pseudo-disk in F. For the family E and an integer k > 0, let E k denote the family of subsets of E of size exactly k, and let E k be the family of subsets of E of size at most k. In this section we bound the number E k of sets of size at most k in the set system E, where k n is an integer parameter. We show that this number is only linear in n and polynomial in k. Moreover, this property is hereditary, that is, it holds for any restriction of E to a subset of P. Then, using the machinery of Chan et al. [5], we conclude that a constant approximation factor for the weighted dominating set problem can be achieved in expected polynomial time. We begin with the case k = 2 and proceed to the cases k = 3, k = 4, and k > 4. In the latter three cases we use a variant of the analysis and some of the observations made by Buzaglo et al. [4], rephrasing them in the context of dominating sets, rather than that of hitting sets, as they were originally formulated. Using the analysis in Section 3, we first show that the number of pairs is linear: Theorem 4. In any range space (P, E), P F, as above, E 2 = O( P ). Proof. Let U be the set of pseudo-disks of P, each of which has at least one point not covered by other pseudo-disks of P, and put C = P \ U. Equivalently, each pseudo-disk of U contains at least one level-1 face of the arrangement A = A(P), while those of C contain only faces of level-2 or higher, as we have assumed general position. By definition, the family U is well behaved. Ranges of E 2 (or pairs) can be classified into those that only contain pseudodisks of U and those that contain at least one pseudo-disk of C. As U is well behaved, the number of pairs of the former kind is O( U ) = O( P ), by Lemma 2. Now consider a pair {D, D } E 2 with D C. It must correspond to some pseudo-disk P F meeting D and D and no other pseudo-disks of P. By the assumption of general position P D must contain a point p in the interior of face f of A contained in D. Since D C, f must be at level 2 (if f is at level above 2, p is contained in more than two pseudo-disks of P, contrary to our assumption). Thus it must be contained in another pseudo-disk of P, necessarily D. Hence fixing f A determines the pair {D, D } E 2 uniquely and the number of such faces f is at most O( P ), as follows by the discussion about arrangements and levels in Section 2. This concludes the proof of the theorem. 4.1 Bounding the number of triples and quadruples We next show:
10 Theorem 5. In any set system (P, E), P F as above, we have (i) E 3 = O( P ), and (ii) E 4 = O( P ). In order to prove Theorem 5, we first need the following key lemma: Lemma 3. Let k 2 be a fixed integer. Let F be a family of pseudo-disks in the plane. Let H be a subfamily of m pseudo-disks from F. We call a pair of pseudo-disks {D 1, D 2 } from H k-good if there exists a pseudo-disk in F that intersects D 1, D 2, and at most k 2 additional pseudo-disks from H, for a total of most k pseudo-disks from H. Then the number of k-good pairs in H is at most c k m, where c k is an absolute constant depending only on k. Proof. We prove the lemma by induction on k. Case k = 2 is precisely Theorem 4. Suppose k 3. We choose each pseudo-disk in H independently with probability p = 1/2 (but we keep F intact). We denote the resulting sample of pseudo-disks by H. We say that a k-good pair {D 1, D 2 } from H survives if D 1, D 2 H and there is a pseudo-disk in F that intersects D 1, D 2, and a total of at most k 1 pseudo-disks in H. In other words, after sampling {D 1, D 2 } becomes (k 1)-good. We observe that a k-good pair {D 1, D 2 } in H survives with probability of at least 1/8. Indeed, because {D 1, D 2 } is a k-good pair, there exists F F such that F intersects D 1 and D 2 and a total of l k pseudo-disks in H. If l k 1, then {D 1, D 2 } is (k 1)-good as soon as both D 1 and D 2 are in H ; this happens with probability 1/4. If l = k, let S H be a pseudo-disk other than D 1 and D 2 intersected by F. If D 1 and D 2 are in H and S is not in H, then {D 1, D 2 } becomes (k 1)-good. This happens with probability 1/8; there may be other ways for {D 1, D 2 } to become (k 1)-good. Therefore the expected number of (k 1)-good pairs in H is at least 1 8 of the number of k-good pairs in H. By the inductive hypothesis on H, there are at most c k 1 H (k 1)-good pairs of disks in H. Therefore, the expected number of (k 1)-good pairs of disks in H is at most c k 1 m/2. Combining the two estimates, the number of k-good pairs in H is at most 4c k 1 m, as claimed. Theorem 5 is then proved using the following result from [4] (see Appendix C): Lemma 4 (Buzaglo et al. [4]). Let H be a graph on m vertices, with the property that, in any subgraph induced by a subset V of vertices, the number of edges is at most c V, where c > 0 is an absolute constant. Then, for any k 2, the number of copies of K k (the complete graph on k vertices) in H is at most d k m, where d k = (2c)k 1 k!. 4.2 Wrapping up Next we would like to bound the number of sets in E of cardinality at most k, for k 5. We show below that the number of such sets is O( P k 3 ). We could easily generalize Theorem 5 to every k and get a linear bound in P for the number of such sets at the cost of a multiplicative constant that grows extremely fast
11 (super exponentially) in k. In order to overcome this problem and improve the dependence on k, we use Theorem 5 and a fundamental property shown in [4], namely, that in a set system of bounded VC-dimension every set has a unique small signature. Specifically: Theorem 6 (Buzaglo et al. [4]). Let S = {S 1,..., S m } be a set family with VC-dimension d. Then it is possible to assign to each set S S a subset S S (its signature), of cardinality at most d, such that different sets S, S S are assigned distinct signatures. Given this machinery we conclude (see Appendix C for the proof): Proposition 1. Let F be a family of pseudo-disks in the plane and let P F be a finite subfamily. Consider the set system (P, E), as above. Then the number of sets in E of cardinality at most k is O(nk 3 ), for any k 2, where n = P. Remark: From the considerations made by our analysis (see also Appendix C), it is easy to verify that the bound in Proposition 1 is hereditary, that is, for any subcollection P P of m n pseudo-disks, the number of sets in the restriction of E to P, of cardinality at most k, is O(mk 3 ). 5 Concluding Remarks and Open Problems In this paper we have presented an expected polynomial-time algorithm, which produces the smallest weight dominating set in a collection of pseudo-disks in the plane, up to a constant factor. Note that our construction of the dominating set does not exploit the geometry of the input objects. In fact, one only needs a predicate that tests whether a pair of pseudo-disks intersect, in order to produce the incidence matrix A. The main open problem raised by our study is whether one can obtain a PTAS, such as the one shown by Gibson and Pirwani [15] for the case of Euclidean disks. References 1. P. K. Agarwal and M. Sharir. Arrangements and their applications. Handbook of Computational Geometry. (J. Sack and J. Urrutia, eds.), Elsevier, Amsterdam, 2000, B. Aronov, E. Ezra, and M. Sharir, Small-size ε-nets for axis-parallel rectangles and boxes. SIAM J. Comput., 39 (2010), H. Brönnimann and M. T. Goodrich. Almost optimal set covers in finite VC dimensions. Discrete Comput. Geom., 14 (1995), S. Buzaglo, R. Pinchasi, and G. Rote. Topological Hyper-graphs. In J. Pach (ed.), Thirty Essays on Geometric Graph Theory, pp , T. Chan, E. Grant, J. Koenemann, and M. Sharpe. Weighted capacitated, priority, and geometric set cover via improved quasi-uniform sampling. In Proc. Symp. Discrete Algorithms (SODA 2012), pp , 2012.
12 6. M. Chlebík and J. Chlebíková. Approximation hardness of dominating set problems. In Proc. 12th Annu. Europ. Sympos. Algorithms (ESA 2004), pp , V. Chvátal. A greedy heuristic for the set-covering problem. Mathematics of Operations Research, 4(3): , B. N. Clark, C. J. Colbourn, and D. S. Johnson. Unit disk graphs. Discrete Mathematics, 86(1-3): , K. L. Clarkson and P. W. Shor. Applications of random sampling in computational geometry, II. Discrete Comput. Geom., 4 (1989), I. Dinur and D. Steurer. Analytical approach to parallel repetition. In Proc. 46th Annu. ACM Symp. Theory Comput., pp , T. Erlebach and E. J. van Leeuwen. Domination in geometric intersection graphs. In LATIN, pages , T. Erlebach and M. Mihalák. A (4 + ε)-approximation for the minimum-weight dominating set problem in unit disk graphs. In Proc. 7th Workshop Approx. and Online Algorithms (WAOA 2009), Revised Papers. LNCS 5893, Springer-Verlag, pp , G. Even, D. Rawitz, and S. Shahar. Hitting sets when the VC-dimension is small. Inform. Process. Letts., 95 (2005), M. R. Garey and D. S. Johnson. Computers and Intractability: A Guide to the Theory of NP-Completeness. W. H. Freeman, New York, NY, M. Gibson and I. A. Pirwani. Approximation algorithms for dominating set in disk graphs: Breaking the log n barrier. Proc. 18th Annu. Europ. Symp. Algorithms, 2010, S. Har-Peled. Geometric Approximation Algorithms. Mathematical Surveys and Monographs, Vol. 173, S. Har-Peled, and K. Quanrud. Approximation algorithms for polynomial-expansion and low-density graphs. In Proc. 23rd Annual European Symposium, pp , H. B. Hunt III, M. V. Marathe, V. Radhakrishnan, S. S. Ravi, D. J. Rosenkrantz, and R. E. Stearns. NC-approximation schemes for NP- and PSPACE-hard problems for geometric graphs. J. Algorithms, 26(2): , L. Lovász. On the ratio of optimal integral and fractional covers. Discrete Mathematics, 13: , R. M. Karp. Reducibility among combinatorial problems. In R. E. Miller, J. W. Thatcher, Eds., Complexity of Computer Computations, pages , Plenum Press, New York, K. Kedem, R. Livne, J. Pach, and M. Sharir. On the union of Jordan regions and collision-free translational motion amidst polygonal obstacles. Discrete Comput. Geom., 1 (1986), J. Matoušek. Approximations and optimal geometric divide-and-conquer. J. Comput Sys. Sci., 50 (1995), pp T. Nieberg, and J. L. Hurink. A PTAS for the minimum dominating set problem in unit disk graphs. Proc. Third Workshop Approx. and Online Algorithms (WAOA 2005), Revised Papers, LNCS 3879, Springer-Verlag, pp , M.J. Pelsmajer, M. Schaefer, D. Štefankovič. Removing even crossings. Journal of Combinatorial Theory, Ser. B, 97(4): , W. T. Tutte. Toward a theory of crossing numbers. J. Combinat. Theory, 8:45 53, K. Varadarajan. Weighted geometric set cover via quasi-uniform sampling. In Proc. 42th Annu. ACM Symp. Theory Comput., pp , 2010.
13 Appendix A Undoing Self-intersections Fig. 2. How to undo self-intersections B Shattering Four Pseudo-disks Fig. 3. How to shatter a set of four shaded pseudo-disk objects: the pseudo-disks meeting all or none of four objects are not shown. The pseudo-disks meeting exactly one object are the objects themselves. C Omitted Proofs from Section 4 Proof of Theorem 5 (i): We follow the approach in [4]. We define a graph G whose vertices are the pseudo-disks in P. Two pseudo-disks in P form an edge in G if they belong to some set in E of cardinality 3.
14 We claim that, if G is an induced subgraph of G, then the number of edges in G is O( V (G ) ), where V (G ) is the set of vertices of G. To see this we only need to apply Lemma 3 with k = 3 and H = V (G ). We can now use Lemma 4 and conclude that the number of triangles in G is at most linear in V (G). This is sufficient to prove the assertion in Theorem 5, since every triple {D 1, D 2, D 3 } E gives rise to a unique triangle in G. Proof of Theorem 5 (ii): We repeat the proof of part (i) almost verbatim. We define a graph G whose vertices are the pseudo-disks in P. Two pseudo-disks in P form an edge in G if they belong to some set in E of cardinality 4. We claim that, if G is an induced subgraph of G, then the number of edges in G is O( V (G ) ). This follows from Lemma 3 with k = 4 and H = V (G ). We can now apply Lemma 4 and conclude that the number of copies of K 4 in G is at most linear in V (G). This completes the proof of the theorem, as is easily verified. Proof of Proposition 1: We follow almost verbatim the random sampling approach in [4]. Using Theorem 3, the VC-dimension of (P, E) is at most 4. Applying Theorem 6, we assign to each S E a unique subset B S S of cardinality at most 4. Let 0 < q < 1/2 be a parameter to be fixed shortly. We now select each pseudo-disk in P independently with probability q. Let P be the resulting sample, and consider the set system E restricted to P, denote it by (P, E ). We say that S E survives if all the disks in B S are in P but none of the remaining disks in S \ B S are in P. It is easy to verify that, if a set S of E has cardinality at most k, then S survives with probability q B S (1 q) S B S q B S (1 q) k B S q 4 (1 q) k 4, where the first inequality follows from the assumption S k, and the second from the fact that q < 1/2. By Theorems 4 and 5, the number of sets in E of cardinality 2, 3, and 4 is O( P ). Clearly, the number of sets in E of cardinality 1 is at most P. It thus follows that the number of surviving sets S from (P, E) is O( P ). Taking expectations, we see that the expected number of surviving sets S from E is O( P ) = O(q P ). On the other hand, the expected number of surviving sets S from E with cardinality at most k is greater than or equal to q 4 (1 q) k 4 E k (recall that E k is the number of sets in E with cardinality at most k). Therefore, q 4 (1 q) k 4 E k = O(q P ). By setting q = 1/k, we obtain E k = O( P k 3 ), as asserted.
A Note on Smaller Fractional Helly Numbers
A Note on Smaller Fractional Helly Numbers Rom Pinchasi October 4, 204 Abstract Let F be a family of geometric objects in R d such that the complexity (number of faces of all dimensions on the boundary
More informationOn Regular Vertices of the Union of Planar Convex Objects
On Regular Vertices of the Union of Planar Convex Objects Esther Ezra János Pach Micha Sharir Abstract Let C be a collection of n compact convex sets in the plane, such that the boundaries of any pair
More informationApproximation Algorithms for Dominating Set in Disk Graphs
Approximation Algorithms for Dominating Set in Disk Graphs Matt Gibson 1 and Imran A. Pirwani 2 1 Dept. of Computer Science, University of Iowa, Iowa City, IA 52242, USA. mrgibson@cs.uiowa.edu 2 Dept.
More informationA non-linear lower bound for planar epsilon-nets
A non-linear lower bound for planar epsilon-nets Noga Alon Abstract We show that the minimum possible size of an ɛ-net for point objects and line (or rectangle)- ranges in the plane is (slightly) bigger
More informationA CONSTANT-FACTOR APPROXIMATION FOR MULTI-COVERING WITH DISKS
A CONSTANT-FACTOR APPROXIMATION FOR MULTI-COVERING WITH DISKS Santanu Bhowmick, Kasturi Varadarajan, and Shi-Ke Xue Abstract. We consider the following multi-covering problem with disks. We are given two
More informationCutting Circles into Pseudo-segments and Improved Bounds for Incidences
Cutting Circles into Pseudo-segments and Improved Bounds for Incidences Boris Aronov Micha Sharir December 11, 2000 Abstract We show that n arbitrary circles in the plane can be cut into On 3/2+ε arcs,
More informationA Simple Proof of Optimal Epsilon Nets
A Simple Proof of Optimal Epsilon Nets Nabil H. Mustafa Kunal Dutta Arijit Ghosh Abstract Showing the existence of ɛ-nets of small size has been the subject of investigation for almost 30 years, starting
More informationLinear-Time Approximation Algorithms for Unit Disk Graphs
Linear-Time Approximation Algorithms for Unit Disk Graphs Guilherme D. da Fonseca Vinícius G. Pereira de Sá elina M. H. de Figueiredo Universidade Federal do Rio de Janeiro, Brazil Abstract Numerous approximation
More informationSmall-size ε-nets for Axis-Parallel Rectangles and Boxes
Small-size ε-nets for Axis-Parallel Rectangles and Boxes Boris Aronov Esther Ezra Micha Sharir March 28, 2009 Abstract We show the existence of ε-nets of size O ( 1 ε log log ε) 1 for planar point sets
More information4 CONNECTED PROJECTIVE-PLANAR GRAPHS ARE HAMILTONIAN. Robin Thomas* Xingxing Yu**
4 CONNECTED PROJECTIVE-PLANAR GRAPHS ARE HAMILTONIAN Robin Thomas* Xingxing Yu** School of Mathematics Georgia Institute of Technology Atlanta, Georgia 30332, USA May 1991, revised 23 October 1993. Published
More informationBichain graphs: geometric model and universal graphs
Bichain graphs: geometric model and universal graphs Robert Brignall a,1, Vadim V. Lozin b,, Juraj Stacho b, a Department of Mathematics and Statistics, The Open University, Milton Keynes MK7 6AA, United
More informationTesting Equality in Communication Graphs
Electronic Colloquium on Computational Complexity, Report No. 86 (2016) Testing Equality in Communication Graphs Noga Alon Klim Efremenko Benny Sudakov Abstract Let G = (V, E) be a connected undirected
More informationAn algorithm to increase the node-connectivity of a digraph by one
Discrete Optimization 5 (2008) 677 684 Contents lists available at ScienceDirect Discrete Optimization journal homepage: www.elsevier.com/locate/disopt An algorithm to increase the node-connectivity of
More informationA better approximation ratio for the Vertex Cover problem
A better approximation ratio for the Vertex Cover problem George Karakostas Dept. of Computing and Software McMaster University October 5, 004 Abstract We reduce the approximation factor for Vertex Cover
More informationSergey Norin Department of Mathematics and Statistics McGill University Montreal, Quebec H3A 2K6, Canada. and
NON-PLANAR EXTENSIONS OF SUBDIVISIONS OF PLANAR GRAPHS Sergey Norin Department of Mathematics and Statistics McGill University Montreal, Quebec H3A 2K6, Canada and Robin Thomas 1 School of Mathematics
More informationSUB-EXPONENTIALLY MANY 3-COLORINGS OF TRIANGLE-FREE PLANAR GRAPHS
SUB-EXPONENTIALLY MANY 3-COLORINGS OF TRIANGLE-FREE PLANAR GRAPHS Arash Asadi Luke Postle 1 Robin Thomas 2 School of Mathematics Georgia Institute of Technology Atlanta, Georgia 30332-0160, USA ABSTRACT
More informationPolynomial time approximation schemes for minimum disk cover problems
DOI 10.1007/s10878-009-9216-y Polynomial time approximation schemes for minimum disk cover problems Chen Liao Shiyan Hu Springer Science+Business Media, LLC 2009 Abstract The following planar minimum disk
More informationBounds on parameters of minimally non-linear patterns
Bounds on parameters of minimally non-linear patterns P.A. CrowdMath Department of Mathematics Massachusetts Institute of Technology Massachusetts, U.S.A. crowdmath@artofproblemsolving.com Submitted: Jan
More informationarxiv: v1 [cs.cg] 16 May 2011
Collinearities in Kinetic Point Sets arxiv:1105.3078v1 [cs.cg] 16 May 2011 Ben D. Lund George B. Purdy Justin W. Smith Csaba D. Tóth August 24, 2018 Abstract Let P be a set of n points in the plane, each
More informationTight Lower Bounds for the Size of Epsilon-Nets
Tight Lower Bounds for the Size of Epsilon-Nets [Extended Abstract] ABSTRACT János Pach EPFL, Lausanne and Rényi Institute, Budapest pach@cims.nyu.edu According to a well known theorem of Haussler and
More informationLecture 9: Random sampling, ɛ-approximation and ɛ-nets
CPS296.2 Geometric Optimization February 13, 2007 Lecture 9: Random sampling, -approximation and -nets Lecturer: Pankaj K. Agarwal Scribe: Shashidhara K. Ganjugunte In the next few lectures we will be
More informationFly Cheaply: On the Minimum Fuel Consumption Problem
Journal of Algorithms 41, 330 337 (2001) doi:10.1006/jagm.2001.1189, available online at http://www.idealibrary.com on Fly Cheaply: On the Minimum Fuel Consumption Problem Timothy M. Chan Department of
More informationGeneralized Pigeonhole Properties of Graphs and Oriented Graphs
Europ. J. Combinatorics (2002) 23, 257 274 doi:10.1006/eujc.2002.0574 Available online at http://www.idealibrary.com on Generalized Pigeonhole Properties of Graphs and Oriented Graphs ANTHONY BONATO, PETER
More informationThe Maximum Flow Problem with Disjunctive Constraints
The Maximum Flow Problem with Disjunctive Constraints Ulrich Pferschy Joachim Schauer Abstract We study the maximum flow problem subject to binary disjunctive constraints in a directed graph: A negative
More informationPartial characterizations of clique-perfect graphs II: diamond-free and Helly circular-arc graphs
Partial characterizations of clique-perfect graphs II: diamond-free and Helly circular-arc graphs Flavia Bonomo a,1, Maria Chudnovsky b,2 and Guillermo Durán c,3 a Departamento de Matemática, Facultad
More informationGeometry. The k Most Frequent Distances in the Plane. József Solymosi, 1 Gábor Tardos, 2 and Csaba D. Tóth Introduction
Discrete Comput Geom 8:639 648 (00 DOI: 10.1007/s00454-00-896-z Discrete & Computational Geometry 00 Springer-Verlag New York Inc. The k Most Frequent Distances in the Plane József Solymosi, 1 Gábor Tardos,
More informationCOMPLEXITY OF SHORT RECTANGLES AND PERIODICITY
COMPLEXITY OF SHORT RECTANGLES AND PERIODICITY VAN CYR AND BRYNA KRA Abstract. The Morse-Hedlund Theorem states that a bi-infinite sequence η in a finite alphabet is periodic if and only if there exists
More informationThe maximum edge-disjoint paths problem in complete graphs
Theoretical Computer Science 399 (2008) 128 140 www.elsevier.com/locate/tcs The maximum edge-disjoint paths problem in complete graphs Adrian Kosowski Department of Algorithms and System Modeling, Gdańsk
More informationMinimum-Dilation Tour (and Path) is NP-hard
Minimum-Dilation Tour (and Path) is NP-hard Panos Giannopoulos Christian Knauer Dániel Marx Abstract We prove that computing a minimum-dilation (Euclidean) Hamilton circuit or path on a given set of points
More informationMaximum union-free subfamilies
Maximum union-free subfamilies Jacob Fox Choongbum Lee Benny Sudakov Abstract An old problem of Moser asks: how large of a union-free subfamily does every family of m sets have? A family of sets is called
More informationOut-colourings of Digraphs
Out-colourings of Digraphs N. Alon J. Bang-Jensen S. Bessy July 13, 2017 Abstract We study vertex colourings of digraphs so that no out-neighbourhood is monochromatic and call such a colouring an out-colouring.
More informationPerfect matchings in highly cyclically connected regular graphs
Perfect matchings in highly cyclically connected regular graphs arxiv:1709.08891v1 [math.co] 6 Sep 017 Robert Lukot ka Comenius University, Bratislava lukotka@dcs.fmph.uniba.sk Edita Rollová University
More informationThe Turán number of sparse spanning graphs
The Turán number of sparse spanning graphs Noga Alon Raphael Yuster Abstract For a graph H, the extremal number ex(n, H) is the maximum number of edges in a graph of order n not containing a subgraph isomorphic
More informationEfficient independent set approximation in unit disk graphs
Efficient independent set approximation in unit disk graphs Gautam Das, Guilherme Da Fonseca, Ramesh Jallu To cite this version: Gautam Das, Guilherme Da Fonseca, Ramesh Jallu. Efficient independent set
More informationEuclidean Shortest Path Algorithm for Spherical Obstacles
Euclidean Shortest Path Algorithm for Spherical Obstacles Fajie Li 1, Gisela Klette 2, and Reinhard Klette 3 1 College of Computer Science and Technology, Huaqiao University Xiamen, Fujian, China 2 School
More informationConflict-Free Colorings of Rectangles Ranges
Conflict-Free Colorings of Rectangles Ranges Khaled Elbassioni Nabil H. Mustafa Max-Planck-Institut für Informatik, Saarbrücken, Germany felbassio, nmustafag@mpi-sb.mpg.de Abstract. Given the range space
More informationThe minimum G c cut problem
The minimum G c cut problem Abstract In this paper we define and study the G c -cut problem. Given a complete undirected graph G = (V ; E) with V = n, edge weighted by w(v i, v j ) 0 and an undirected
More informationFast algorithms for even/odd minimum cuts and generalizations
Fast algorithms for even/odd minimum cuts and generalizations András A. Benczúr 1 and Ottilia Fülöp 2 {benczur,otti}@cs.elte.hu 1 Computer and Automation Institute, Hungarian Academy of Sciences, and Department
More informationGraph coloring, perfect graphs
Lecture 5 (05.04.2013) Graph coloring, perfect graphs Scribe: Tomasz Kociumaka Lecturer: Marcin Pilipczuk 1 Introduction to graph coloring Definition 1. Let G be a simple undirected graph and k a positive
More informationComputing a Minimum-Dilation Spanning Tree is NP-hard
Computing a Minimum-Dilation Spanning Tree is NP-hard Otfried Cheong 1 Herman Haverkort 2 Mira Lee 1 October 0, 2008 Abstract In a geometric network G = (S, E), the graph distance between two vertices
More informationGraph Classes and Ramsey Numbers
Graph Classes and Ramsey Numbers Rémy Belmonte, Pinar Heggernes, Pim van t Hof, Arash Rafiey, and Reza Saei Department of Informatics, University of Bergen, Norway Abstract. For a graph class G and any
More informationModel-theoretic distality and incidence combinatorics
Model-theoretic distality and incidence combinatorics Artem Chernikov UCLA Model Theory and Combinatorics workshop IHP, Paris, France Jan 29, 2018 Intro I will discuss generalizations of various results
More informationExcluded permutation matrices and the Stanley Wilf conjecture
Excluded permutation matrices and the Stanley Wilf conjecture Adam Marcus Gábor Tardos November 2003 Abstract This paper examines the extremal problem of how many 1-entries an n n 0 1 matrix can have that
More informationDispersing Points on Intervals
Dispersing Points on Intervals Shimin Li 1 and Haitao Wang 1 Department of Computer Science, Utah State University, Logan, UT 843, USA shiminli@aggiemail.usu.edu Department of Computer Science, Utah State
More informationShortest paths with negative lengths
Chapter 8 Shortest paths with negative lengths In this chapter we give a linear-space, nearly linear-time algorithm that, given a directed planar graph G with real positive and negative lengths, but no
More informationCovering the Convex Quadrilaterals of Point Sets
Covering the Convex Quadrilaterals of Point Sets Toshinori Sakai, Jorge Urrutia 2 Research Institute of Educational Development, Tokai University, 2-28-4 Tomigaya, Shibuyaku, Tokyo 5-8677, Japan 2 Instituto
More informationc 2010 Society for Industrial and Applied Mathematics
SIAM J. DISCRETE MATH. Vol. 24, No. 3, pp. 1038 1045 c 2010 Society for Industrial and Applied Mathematics SET SYSTEMS WITHOUT A STRONG SIMPLEX TAO JIANG, OLEG PIKHURKO, AND ZELEALEM YILMA Abstract. A
More informationVoronoi Diagrams for Oriented Spheres
Voronoi Diagrams for Oriented Spheres F. Aurenhammer J. Wallner University of Technology Graz, Austria auren@igi.tugraz.at j.wallner@tugraz.at M. Peternell H. Pottmann University of Technology Vienna,
More informationSize and degree anti-ramsey numbers
Size and degree anti-ramsey numbers Noga Alon Abstract A copy of a graph H in an edge colored graph G is called rainbow if all edges of H have distinct colors. The size anti-ramsey number of H, denoted
More informationImproved Bounds on the Union Complexity of Fat Objects
Discrete Comput Geom (2008) 40: 127 140 DOI 10.1007/s00454-007-9029-7 Improved Bounds on the Union Complexity of Fat Objects Mark de Berg Received: 20 October 2006 / Revised: 9 July 2007 / Published online:
More information18.5 Crossings and incidences
18.5 Crossings and incidences 257 The celebrated theorem due to P. Turán (1941) states: if a graph G has n vertices and has no k-clique then it has at most (1 1/(k 1)) n 2 /2 edges (see Theorem 4.8). Its
More informationOn Dominator Colorings in Graphs
On Dominator Colorings in Graphs Ralucca Michelle Gera Department of Applied Mathematics Naval Postgraduate School Monterey, CA 994, USA ABSTRACT Given a graph G, the dominator coloring problem seeks a
More informationConflict-Free Coloring and its Applications
Conflict-Free Coloring and its Applications Shakhar Smorodinsky September 12, 2011 Abstract Let H = (V, E) be a hypergraph. A conflict-free coloring of H is an assignment of colors to V such that, in each
More informationVertex colorings of graphs without short odd cycles
Vertex colorings of graphs without short odd cycles Andrzej Dudek and Reshma Ramadurai Department of Mathematical Sciences Carnegie Mellon University Pittsburgh, PA 1513, USA {adudek,rramadur}@andrew.cmu.edu
More informationSmall-size ε-nets for Axis-Parallel Rectangles and Boxes
Small-size ε-nets for Axis-Parallel Rectangles and Boxes [Extended Abstract] Boris Aronov Department of Computer Science and Engineering Polytechnic Institute of NYU Brooklyn, NY 11201-3840, USA aronov@poly.edu
More informationAhlswede Khachatrian Theorems: Weighted, Infinite, and Hamming
Ahlswede Khachatrian Theorems: Weighted, Infinite, and Hamming Yuval Filmus April 4, 2017 Abstract The seminal complete intersection theorem of Ahlswede and Khachatrian gives the maximum cardinality of
More informationSpanning Paths in Infinite Planar Graphs
Spanning Paths in Infinite Planar Graphs Nathaniel Dean AT&T, ROOM 2C-415 600 MOUNTAIN AVENUE MURRAY HILL, NEW JERSEY 07974-0636, USA Robin Thomas* Xingxing Yu SCHOOL OF MATHEMATICS GEORGIA INSTITUTE OF
More informationProbabilistic Proofs of Existence of Rare Events. Noga Alon
Probabilistic Proofs of Existence of Rare Events Noga Alon Department of Mathematics Sackler Faculty of Exact Sciences Tel Aviv University Ramat-Aviv, Tel Aviv 69978 ISRAEL 1. The Local Lemma In a typical
More informationLaplacian Integral Graphs with Maximum Degree 3
Laplacian Integral Graphs with Maximum Degree Steve Kirkland Department of Mathematics and Statistics University of Regina Regina, Saskatchewan, Canada S4S 0A kirkland@math.uregina.ca Submitted: Nov 5,
More informationScheduling Parallel Jobs with Linear Speedup
Scheduling Parallel Jobs with Linear Speedup Alexander Grigoriev and Marc Uetz Maastricht University, Quantitative Economics, P.O.Box 616, 6200 MD Maastricht, The Netherlands. Email: {a.grigoriev, m.uetz}@ke.unimaas.nl
More informationWeighted Capacitated, Priority, and Geometric Set Cover via Improved Quasi-Uniform Sampling
Weighted Capacitated, Priority, and Geometric Set Cover via Improved Quasi-Uniform Sampling Timothy M. Chan Elyot Grant. Jochen Könemann Malcolm Sharpe October 6, 2011 Abstract The minimum-weight set cover
More informationPacking and Covering Dense Graphs
Packing and Covering Dense Graphs Noga Alon Yair Caro Raphael Yuster Abstract Let d be a positive integer. A graph G is called d-divisible if d divides the degree of each vertex of G. G is called nowhere
More informationAn approximate version of Hadwiger s conjecture for claw-free graphs
An approximate version of Hadwiger s conjecture for claw-free graphs Maria Chudnovsky Columbia University, New York, NY 10027, USA and Alexandra Ovetsky Fradkin Princeton University, Princeton, NJ 08544,
More informationDistinct Distances in Three and Higher Dimensions
Distinct Distances in Three and Higher Dimensions Boris Aronov János Pach Micha Sharir Gábor Tardos Abstract Improving an old result of Clarkson et al., we show that the number of distinct distances determined
More informationDistinct Distances in Three and Higher Dimensions
Distinct Distances in Three and Higher Dimensions Boris Aronov János Pach Micha Sharir Gábor Tardos ABSTRACT Improving an old result of Clarkson et al., we show that the number of distinct distances determined
More informationApproximation algorithms and hardness results for the clique packing problem. October, 2007
Approximation algorithms and hardness results for the clique packing problem F. Chataigner 1 G. Manić 2 Y.Wakabayashi 1 R. Yuster 3 1 Instituto de Matemática e Estatística Universidade de São Paulo, SP,
More informationPIERCING AXIS-PARALLEL BOXES
PIERCING AXIS-PARALLEL BOXES MARIA CHUDNOVSKY, SOPHIE SPIRKL, AND SHIRA ZERBIB Abstract. Given a finite family F of axis-parallel boxes in R d such that F contains no k + 1 pairwise disjoint boxes, and
More informationarxiv: v1 [math.co] 25 Apr 2016
On the zone complexity of a ertex Shira Zerbib arxi:604.07268 [math.co] 25 Apr 206 April 26, 206 Abstract Let L be a set of n lines in the real projectie plane in general position. We show that there exists
More informationFirst Order Convergence and Roots
Article First Order Convergence and Roots Christofides, Demetres and Kral, Daniel Available at http://clok.uclan.ac.uk/17855/ Christofides, Demetres and Kral, Daniel (2016) First Order Convergence and
More informationIntroduction to Graph Theory
Introduction to Graph Theory George Voutsadakis 1 1 Mathematics and Computer Science Lake Superior State University LSSU Math 351 George Voutsadakis (LSSU) Introduction to Graph Theory August 2018 1 /
More informationThe Complexity of Maximum. Matroid-Greedoid Intersection and. Weighted Greedoid Maximization
Department of Computer Science Series of Publications C Report C-2004-2 The Complexity of Maximum Matroid-Greedoid Intersection and Weighted Greedoid Maximization Taneli Mielikäinen Esko Ukkonen University
More informationGenerating p-extremal graphs
Generating p-extremal graphs Derrick Stolee Department of Mathematics Department of Computer Science University of Nebraska Lincoln s-dstolee1@math.unl.edu August 2, 2011 Abstract Let f(n, p be the maximum
More informationOn Approximating the Depth and Related Problems
On Approximating the Depth and Related Problems Boris Aronov Polytechnic University, Brooklyn, NY Sariel Har-Peled UIUC, Urbana, IL 1: Motivation: Operation Inflicting Freedom Input: R - set red points
More informationApproximability of Connected Factors
Approximability of Connected Factors Kamiel Cornelissen 1, Ruben Hoeksma 1, Bodo Manthey 1, N. S. Narayanaswamy 2, and C. S. Rahul 2 1 University of Twente, Enschede, The Netherlands {k.cornelissen, r.p.hoeksma,
More informationAn Improved Approximation Algorithm for Maximum Edge 2-Coloring in Simple Graphs
An Improved Approximation Algorithm for Maximum Edge 2-Coloring in Simple Graphs Zhi-Zhong Chen Ruka Tanahashi Lusheng Wang Abstract We present a polynomial-time approximation algorithm for legally coloring
More informationTheorem (Special Case of Ramsey s Theorem) R(k, l) is finite. Furthermore, it satisfies,
Math 16A Notes, Wee 6 Scribe: Jesse Benavides Disclaimer: These notes are not nearly as polished (and quite possibly not nearly as correct) as a published paper. Please use them at your own ris. 1. Ramsey
More informationOn shredders and vertex connectivity augmentation
On shredders and vertex connectivity augmentation Gilad Liberman The Open University of Israel giladliberman@gmail.com Zeev Nutov The Open University of Israel nutov@openu.ac.il Abstract We consider the
More informationUnion of Random Minkowski Sums and Network Vulnerability Analysis
Union of Random Minkowski Sums and Network Vulnerability Analysis Pankaj K. Agarwal Sariel Har-Peled Haim Kaplan Micha Sharir June 29, 2014 Abstract Let C = {C 1,..., C n } be a set of n pairwise-disjoint
More informationGeometry. Open Caps and Cups in Planar Point Sets. Pavel Valtr. 1. Introduction
Discrete Comput Geom 37:565 576 (2007) DOI: 10.1007/s00454-006-1286-3 Discrete & Computational Geometry 2007 Springer Science+Business Media, Inc. Open Caps and Cups in Planar Point Sets Pavel Valtr Department
More informationGraphs with large maximum degree containing no odd cycles of a given length
Graphs with large maximum degree containing no odd cycles of a given length Paul Balister Béla Bollobás Oliver Riordan Richard H. Schelp October 7, 2002 Abstract Let us write f(n, ; C 2k+1 ) for the maximal
More informationApproximation Algorithms for Re-optimization
Approximation Algorithms for Re-optimization DRAFT PLEASE DO NOT CITE Dean Alderucci Table of Contents 1.Introduction... 2 2.Overview of the Current State of Re-Optimization Research... 3 2.1.General Results
More informationUNAVOIDABLE INDUCED SUBGRAPHS IN LARGE GRAPHS WITH NO HOMOGENEOUS SETS
UNAVOIDABLE INDUCED SUBGRAPHS IN LARGE GRAPHS WITH NO HOMOGENEOUS SETS MARIA CHUDNOVSKY, RINGI KIM, SANG-IL OUM, AND PAUL SEYMOUR Abstract. An n-vertex graph is prime if it has no homogeneous set, that
More informationPart V. 17 Introduction: What are measures and why measurable sets. Lebesgue Integration Theory
Part V 7 Introduction: What are measures and why measurable sets Lebesgue Integration Theory Definition 7. (Preliminary). A measure on a set is a function :2 [ ] such that. () = 2. If { } = is a finite
More informationDiscrete Geometry. Problem 1. Austin Mohr. April 26, 2012
Discrete Geometry Austin Mohr April 26, 2012 Problem 1 Theorem 1 (Linear Programming Duality). Suppose x, y, b, c R n and A R n n, Ax b, x 0, A T y c, and y 0. If x maximizes c T x and y minimizes b T
More informationUnsplittable coverings in the plane
Unsplittable coverings in the plane János Pach Dömötör Pálvölgyi May 26, 2015 Abstract Et tu mi li, Brute? (Julius Caesar) A system of sets forms an m-fold covering of a set X if every point of X belongs
More informationOn Lines and Joints. August 17, Abstract
On Lines and Joints Haim Kaplan Micha Sharir Eugenii Shustin August 17, 2009 Abstract Let L be a set of n lines in R d, for d 3. A joint of L is a point incident to at least d lines of L, not all in a
More informationNested Cycles in Large Triangulations and Crossing-Critical Graphs
Nested Cycles in Large Triangulations and Crossing-Critical Graphs César Hernández Vélez 1, Gelasio Salazar,1, and Robin Thomas,2 1 Instituto de Física, Universidad Autónoma de San Luis Potosí. San Luis
More informationCYCLICALLY FIVE CONNECTED CUBIC GRAPHS. Neil Robertson 1 Department of Mathematics Ohio State University 231 W. 18th Ave. Columbus, Ohio 43210, USA
CYCLICALLY FIVE CONNECTED CUBIC GRAPHS Neil Robertson 1 Department of Mathematics Ohio State University 231 W. 18th Ave. Columbus, Ohio 43210, USA P. D. Seymour 2 Department of Mathematics Princeton University
More informationUniformly discrete forests with poor visibility
Uniformly discrete forests with poor visibility Noga Alon August 19, 2017 Abstract We prove that there is a set F in the plane so that the distance between any two points of F is at least 1, and for any
More informationarxiv:cs/ v1 [cs.cg] 7 Feb 2006
Approximate Weighted Farthest Neighbors and Minimum Dilation Stars John Augustine, David Eppstein, and Kevin A. Wortman Computer Science Department University of California, Irvine Irvine, CA 92697, USA
More informationIndependent Transversal Dominating Sets in Graphs: Complexity and Structural Properties
Filomat 30:2 (2016), 293 303 DOI 10.2298/FIL1602293A Published by Faculty of Sciences and Mathematics, University of Niš, Serbia Available at: http://www.pmf.ni.ac.rs/filomat Independent Transversal Dominating
More informationComputing a Minimum-Dilation Spanning Tree is NP-hard
Computing a Minimum-Dilation Spanning Tree is NP-hard Otfried Cheong Herman Haverkort Mira Lee Dept. of Computer Science, Korea Advanced Institute of Science & Technology, Daejeon, South Korea. Email:
More informationDecompositions of graphs into cycles with chords
Decompositions of graphs into cycles with chords Paul Balister Hao Li Richard Schelp May 22, 2017 In memory of Dick Schelp, who passed away shortly after the submission of this paper Abstract We show that
More informationClaw-free Graphs. III. Sparse decomposition
Claw-free Graphs. III. Sparse decomposition Maria Chudnovsky 1 and Paul Seymour Princeton University, Princeton NJ 08544 October 14, 003; revised May 8, 004 1 This research was conducted while the author
More informationRemoval Lemmas with Polynomial Bounds
Removal Lemmas with Polynomial Bounds Lior Gishboliner Asaf Shapira Abstract A common theme in many extremal problems in graph theory is the relation between local and global properties of graphs. One
More informationPacking triangles in regular tournaments
Packing triangles in regular tournaments Raphael Yuster Abstract We prove that a regular tournament with n vertices has more than n2 11.5 (1 o(1)) pairwise arc-disjoint directed triangles. On the other
More informationOn the intersection of infinite matroids
On the intersection of infinite matroids Elad Aigner-Horev Johannes Carmesin Jan-Oliver Fröhlich University of Hamburg 9 July 2012 Abstract We show that the infinite matroid intersection conjecture of
More informationLecture notes on the ellipsoid algorithm
Massachusetts Institute of Technology Handout 1 18.433: Combinatorial Optimization May 14th, 007 Michel X. Goemans Lecture notes on the ellipsoid algorithm The simplex algorithm was the first algorithm
More informationThe Lefthanded Local Lemma characterizes chordal dependency graphs
The Lefthanded Local Lemma characterizes chordal dependency graphs Wesley Pegden March 30, 2012 Abstract Shearer gave a general theorem characterizing the family L of dependency graphs labeled with probabilities
More informationThe τ-skyline for Uncertain Data
CCCG 2014, Halifax, Nova Scotia, August 11 13, 2014 The τ-skyline for Uncertain Data Haitao Wang Wuzhou Zhang Abstract In this paper, we introduce the notion of τ-skyline as an alternative representation
More information