arXiv subjects

Vadim E. Levit

Publications and source records attributed to Vadim E. Levit.

At least 37 records · Page 2Linked to original sources

arXiv2022

Fair Division Algorithms for Electricity Distribution

In many developing countries, the total electricity demand is larger than the limited generation capacity of power stations. Many countries adopt the common practice of routine load shedding - disconnecting entire regions from the power supply - to maintain a balance between demand and supply. Load shedding results in inflicting hardship and discomfort on households, which is even worse and hence unfair to those whose need for electricity is higher than that of others during load shedding hours. Recently, Oluwasuji et al. [2020] presented this problem and suggested several heuristic solutions. In this work, we study the electricity distribution problem as a problem of fair division, model it using the related literature on cake-cutting problems, and discuss some insights on which parts of the time intervals are allocated to each household. We consider four cases: identical demand, uniform utilities; identical demand, additive utilities; different demand, uniform utilities; different demand, additive utilities. We provide the solution for the first two cases and discuss the novel concept of q-times bin packing in relation to the remaining cases. We also show how the fourth case is related to the consensus k-division problem. One can study objectives and constraints using utilitarian and egalitarian social welfare metrics, as well as trying to keep the number of cuts as small as possible. A secondary objective can be to minimize the maximum utility-difference between agents.

cs.GT↗

arXiv2022

A Linear Algorithm for Computing Independence Polynomials of Trees

An independent set in a graph is a set of pairwise non-adjacent vertices. Let $α(G)$ denote the cardinality of a maximum independent set in the graph $G = (V, E)$. Gutman and Harary defined the independence polynomial of $G$ \[ I(G;x) = \sum_{k=0}^{α(G)}{s_k}x^{k}={s_0}+{s_1}x+{s_2}x^{2}+...+{s_{α(G)}}x^{α(G)}, \] where $s_k$ denotes the number of independent sets of cardinality $k$ in the graph $G$. A comprehensive survey on the subject is due to Levit and Mandrescu, where some recursive formulas are allowing to calculate the independence polynomial. A direct implementation of these recursions does not bring about an efficient algorithm. Yosef, Mizrachi, and Kadrawi developed an efficient way for computing the independence polynomials of trees with $n$ vertices, such that a database containing all of the independence polynomials of all the trees with up to $n-1$ vertices is required. This approach is not suitable for big trees, as an extensive database is needed. On the other hand, using dynamic programming, it is possible to develop an efficient algorithm that prevents repeated calculations. In summary, our dynamic programming algorithm runs over a tree in linear time and does not depend on a database.

math.CO↗

arXiv2021

Cospanning characterizations of antimatroids and convex geometries

Given a finite set $E$ and an operator $σ:2^{E}\longrightarrow2^{E}$, two sets $X,Y\subseteq E$ are \textit{cospanning} if $σ\left( X\right) =σ\left( Y\right) $. Corresponding \textit{cospanning equivalence relations} were investigated for greedoids in much detail (Korte, Lovasz, Schrader; 1991). For instance, these relations determine greedoids uniquely. In fact, the feasible sets of a greedoid are exactly the inclusion-wise minimal sets of the equivalence classes. In this research, we show that feasible sets of convex geometries are the inclusion-wise maximal sets of the equivalence classes of the corresponding closure operator. Same as greedoids, convex geometries are uniquely defined by the corresponding cospanning relations. For each closure operator $σ$, an element $x\in X$ is \textit{an extreme point} of $X$ if $x\notinσ(X-x)$. The set of extreme points of $X$ is denoted by $ex(X)$. We prove, that if $σ$ has the anti-exchange property, then for every set $X$ its equivalence class $[X]_σ$ is the interval $[ex(X),σ(X)]$. It results in the one-to-one correspondence between the cospanning partitions of an antimatroid and its complementary convex geometry. The obtained results are based on the connection between violator spaces, greedoids, and antimatroids. Cospanning characterization of these combinatorial structures allows us not only to give the new characterization of antimatroids and convex geometries but also to obtain the new properties of closure operators, extreme point operators, and their interconnections.

math.CO↗

arXiv2021

Cospanning characterizations of violator and co-violator spaces

Given a finite set E and an operator sigma:2^{E}-->2^{E}, two subsets X,Y of the ground set E are cospanning if sigma(X)=sigma(Y) (Korte, Lovasz, Schrader; 1991). We investigate cospanning relations on violator spaces. A notion of a violator space was introduced in (Gartner, Matousek, Rust, Skovrovn; 2008) as a combinatorial framework that encompasses linear programming and other geometric optimization problems. Violator spaces are defined by violator operators. We introduce co-violator spaces based on contracting operators known also as choice functions. Let alpha,beta:2^{E}-->2^{E} be a violator operator and a co-violator operator, respectively. Cospanning characterizations of violator spaces allow us to obtain some new properties of violator operators, co-violator operators, and their interconnections. In particular, we show that uniquely generated violator spaces enjoy so-called Krein-Milman properties, i.e., alpha(beta(X))=alpha(X) and beta(alpha}(X))=beta(X) for every subset X of the ground set E.

math.CO↗

arXiv2020

Signed Hultman Numbers and Signed Generalized Commuting Probability in Finite Groups

Let G be a finite group. Let pi be a permutation from S{n}. We study the distribution of probabilities of equality a{1} a{2} ...a{n-1}a{n}=a{pi{1}}^{epsilon{1}} a{pi_{2}}^{epsilon{2}}...a{pi{n-1}}^{epsilon_{n-1}} a_{pi_{n}}^{epsilon{n}}, when pi varies over all the permutations in S{n}, and epsilon{i} varies over the set {+1, -1}. By the paper "Hultman Numbers and Generalized Commuting Probability in Finite Groups" (2017), The case where all epsilon{i} are +1 led to a close connection to Hultman numbers. In this paper we generalize the results, permitting epsilon{i} to be -1. We describe the spectrum of the probabilities of signed permutation equalities in a finite group G. This spectrum turns out to be closely related to the partition of 2^{n}*n! into a sum of the corresponding signed Hultman numbers.

math.GR↗

arXiv2020

Critical sets, crowns, and local maximum independent sets

A set $S\subseteq V(G)$ is independent (or stable) if no two vertices from $S$ are adjacent, and by $\mathrm{Ind}(G)$ we mean the set of all independent sets of $G$. A set $A\in\mathrm{Ind}(G)$ is critical (and we write $A\in CritIndep(G)$) if $\left\vert A\right\vert -\left\vert N(A)\right\vert =\max\{\left\vert I\right\vert -\left\vert N(I)\right\vert :I\in \mathrm{Ind}(G)\}$, where $N(I)$ denotes the neighborhood of $I$. If $S\in\mathrm{Ind}(G)$ and there is a matching from $N(S)$ into $S$, then $S$ is a crown, and we write $S\in Crown(G)$. Let $Ψ(G)$ be the family of all local maximum independent sets of graph $G$, i.e., $S\inΨ(G)$ if $S$ is a maximum independent set in the subgraph induced by $S\cup N(S)$. In this paper we show that $CritIndep(G)\subseteq Crown(G)$ $\subseteqΨ(G)$ are true for every graph. In addition, we present some classes of graphs where these families coincide and form greedoids or even more general set systems that we call augmentoids.

cs.DM↗

arXiv2019

On the Critical Difference of Almost Bipartite Graphs

A set $S\subseteq V$ is \textit{independent} in a graph $G=\left( V,E\right) $ if no two vertices from $S$ are adjacent. The \textit{independence number} $α(G)$ is the cardinality of a maximum independent set, while $μ(G)$ is the size of a maximum matching in $G$. If $α(G)+μ(G)$ equals the order of $G$, then $G$ is called a \textit{König-Egerváry graph }\cite{dem,ster}. The number $d\left( G\right) =\max\{\left\vert A\right\vert -\left\vert N\left( A\right) \right\vert :A\subseteq V\}$ is called the \textit{critical difference} of $G$ \cite{Zhang} (where $N\left( A\right) =\left\{ v:v\in V,N\left( v\right) \cap A\neq\emptyset\right\} $). It is known that $α(G)-μ(G)\leq d\left( G\right) $ holds for every graph \cite{Levman2011a,Lorentzen1966,Schrijver2003}. In \cite{LevMan5} it was shown that $d(G)=α(G)-μ(G)$ is true for every König-Egerváry graph. A graph $G$ is \textit{(i)} \textit{unicyclic} if it has a unique cycle, \textit{(ii)} \textit{almost bipartite} if it has only one odd cycle. It was conjectured in \cite{LevMan2012a,LevMan2013a} and validated in \cite{Bhattacharya2018} that $d(G)=α(G)-μ(G)$ holds for every unicyclic non-König-Egerváry graph $G$. In this paper we prove that if $G$ is an almost bipartite graph of order $n\left( G\right) $, then $α(G)+μ(G)\in\left\{ n\left( G\right) -1,n\left( G\right) \right\} $. Moreover, for each of these two values, we characterize the corresponding graphs. Further, using these findings, we show that the critical difference of an almost bipartite graph $G$ satisfies \[ d(G)=α(G)-μ(G)=\left\vert \mathrm{core}(G)\right\vert -\left\vert N(\mathrm{core}(G))\right\vert , \] where by \textrm{core}$\left( G\right) $ we mean the intersection of all maximum independent sets.

cs.DM↗

arXiv2019

On an Annihilation Number Conjecture

Let $α(G)$ denote the cardinality of a maximum independent set, while $μ(G)$ be the size of a maximum matching in the graph $G=\left(V,E\right) $. If $α(G)+μ(G)=\left\vert V\right\vert $, then $G$ is a König-Egerváry graph. If $d_{1}\leq d_{2}\leq\cdots\leq d_{n}$ is the degree sequence of $G$, then the annihilation number $h\left(G\right) $ of $G$ is the largest integer $k$ such that $\sum\limits_{i=1}^{k}d_{i}\leq\left\vert E\right\vert $ (Pepper 2004, Pepper 2009). A set $A\subseteq V$ satisfying $\sum \limits_{a\in A} deg(a)\leq\left\vert E\right\vert $ is an annihilation set, if, in addition, $ deg\left(v\right) +\sum\limits_{a\in A} deg(a)>\left\vert E\right\vert $, for every vertex $v\in V(G)-A$, then $A$ is a maximal annihilation set in $G$. In (Larson & Pepper 2011) it was conjectured that the following assertions are equivalent: (i) $α\left(G\right) =h\left(G\right) $; (ii) $G$ is a König-Egerváry graph and every maximum independent set is a maximal annihilating set. In this paper, we prove that the implication "(i) $\Longrightarrow$ (ii)" is correct, while for the opposite direction we provide a series of generic counterexamples. Keywords: maximum independent set, matching, tree, bipartite graph, König-Egerváry graph, annihilation set, annihilation number.

math.CO↗

arXiv2019

On graphs admitting two disjoint maximum independent sets

An independent set A is maximal if it is not a proper subset of an independent set, while A is maximum if it has a maximum size. The problem of whether a graph has a pair of disjoint maximal independent sets was introduced by C. Berge in early 70's. The class of graphs for which every induced subgraph admits two disjoint maximal independent sets was characterized in (Shaudt, 2015). It is known that deciding whether a graph has two disjoint maximal independent sets is a NP-complete problem (Henning et al., 2009). In this paper, we are focused on finding conditions ensuring the existence of two disjoint maximum independent sets.

math.CO↗

arXiv2018

Generating subgraphs in chordal graphs

A graph $G$ is well-covered if all its maximal independent sets are of the same cardinality. Assume that a weight function $w$ is defined on its vertices. Then $G$ is $w$-well-covered if all maximal independent sets are of the same weight. For every graph $G$, the set of weight functions $w$ such that $G$ is $w$-well-covered is a vector space, denoted $WCW(G)$. Let $B$ be a complete bipartite induced subgraph of $G$ on vertex sets of bipartition $B_{X}$ and $B_{Y}$. Then $B$ is generating if there exists an independent set $S$ such that $S \cup B_{X}$ and $S \cup B_{Y}$ are both maximal independent sets of $G$. In the restricted case that a generating subgraph $B$ is isomorphic to $K_{1,1}$, the unique edge in $B$ is called a relating edge. Generating subgraphs play an important role in finding $WCW(G)$. Deciding whether an input graph $G$ is well-covered is co-NP-complete. Hence, finding $WCW(G)$ is co-NP-hard. Deciding whether an edge is relating is NP-complete. Therefore, deciding whether a subgraph is generating is NP-complete as well. A graph is chordal if every induced cycle is a triangle. It is known that finding $WCW(G)$ can be done polynomially in the restricted case that $G$ is chordal. Thus recognizing well-covered chordal graphs is a polynomial problem. We present a polynomial algorithm for recognizing relating edges and generating subgraphs in chordal graphs.

cs.DM↗

arXiv2018

Recognizing generating subgraphs revisited

A graph $G$ is well-covered if all its maximal independent sets are of the same cardinality. Assume that a weight function $w$ is defined on its vertices. Then $G$ is $w$-well-covered if all maximal independent sets are of the same weight. For every graph $G$, the set of weight functions $w$ such that $G$ is $w$-well-covered is a vector space, denoted as $WCW(G).$ Deciding whether an input graph $G$ is well-covered is co-NP-complete. Therefore, finding $WCW(G)$ is co-NP-hard. A generating subgraph of a graph $G$ is an induced complete bipartite subgraph $B$ of $G$ on vertex sets of bipartition $B_{X}$ and $B_{Y}$, such that each of $S \cup B_{X}$ and $S \cup B_{Y}$ is a maximal independent set of $G$, for some independent set $S$. If $B$ is generating, then $w(B_{X})=w(B_{Y})$ for every weight function $w \in WCW(G)$. Therefore, generating subgraphs play an important role in finding $WCW(G)$. The decision problem whether a subgraph of an input graph is generating is known to be NP-complete. In this article, we prove NP-completeness of the problem for graphs without cycles of length 3 and 5, and for bipartite graphs with girth at least 6. On the other and, we supply polynomial algorithms for recognizing generating subgraphs and finding $WCW(G)$, when the input graph is bipartite without cycles of length 6. We also present a polynomial algorithm which finds $WCW(G)$ when $G$ does not contain cycles of lengths 3, 4, 5, and 7.

cs.DM↗

arXiv2016

1-well-covered graphs revisited

A graph is well-covered if all its maximal independent sets are of the same size (M. D. Plummer, 1970). A well-covered graph is 1-well-covered if the deletion of every vertex leaves a graph which is well-covered as well (J. W. Staples, 1975). A graph G belongs to class W_{n} if every n pairwise disjoint independent sets in G are included in $n$ pairwise disjoint maximum independent sets (J. W. Staples, 1975). Clearly, W_{1} is the family of all well-covered graphs. It turns out that G belongs to W_{2} if and only if it is a 1-well-covered graph without isolated vertices. We show that deleting a shedding vertex does not change the maximum size of a maximal independent set including a given independent set A in a graph G. Specifically, for well-covered graphs, it means that the vertex v is shedding if and only if G-v is well-covered. In addition, we provide new characterizations of 1-well-covered graphs, which we further use in building 1-well-covered graphs by corona, join, and concatenation operations.

math.CO↗

arXiv2016

The Roller-Coaster Conjecture Revisited

A graph is well-covered if all its maximal independent sets are of the same cardinality (Plummer, 1970). If G is a well-covered graph, has at least two vertices, and G-v is well-covered for every vertex v, then G is a 1-well-covered graph (Staples, 1975). We call G a λ-quasi-regularizable graph if λ |S| =< |N(S)| for every independent set S of G. The independence polynomial I(G;x) is the generating function of independent sets in a graph G (Gutman & Harary, 1983). The Roller-Coaster Conjecture (Michael & Travis, 2003), saying that for every permutation σ of the set {(α/2),...,α} there exists a well-covered graph G with independence number α such that the coefficients (s_{k}) of I(G;x) are chosen in accordance with σ, has been validated in (Cutler & Pebody, 2017). In this paper, we show that independence polynomials of λ-quasi-regularizable graphs are partially unimodal. More precisely, the coefficients of an upper part of I(G;x) are in non-increasing order. Based on this finding, we prove that the domain of the Roller- Coaster Conjecture can be shortened for well-covered graphs and 1-well-covered graphs.

math.CO↗

arXiv2016

Violator spaces vs closure spaces

Violator Spaces were introduced by J. Matousek et al. in 2008 as generalization of Linear Programming problems. Convex geometries were invented by Edelman and Jamison in 1985 as proper combinatorial abstractions of convexity. Convex geometries are defined by anti-exchange closure operators. We investigate an interrelations between violator spaces and closure spaces and show that violator mapping may be defined by a week version of closure operators. Moreover, we prove that violator spaces with an unique basis satisfies the anti-exchange and the Krein-Milman properties.

math.CO↗

arXiv2016

Complexity results for generating subgraphs

A graph G is well-covered if all its maximal independent sets are of the same cardinality. Assume that a weight function w is defined on its vertices. Then G is w-well-covered if all maximal independent sets are of the same weight. For every graph G, the set of weight functions w such that G is w-well-covered is a vector space, denoted WCW(G). Let B be a complete bipartite induced subgraph of G on vertex sets of bipartition B_X and B_Y. Then B is generating if there exists an independent set S such that S \cup B_X and S \cup B_Y are both maximal independent sets of G. A relating edge is a generating subgraph in the restricted case that B = K_{1,1}. Deciding whether an input graph G is well-covered is co-NP-complete. Therefore finding WCW(G) is co-NP-hard. Deciding whether an edge is relating is co-NP-complete. Therefore, deciding whether a subgraph is generating is co-NP-complete as well. In this article we discuss the connections among these problems, provide proofs for NP-completeness for several restricted cases, and present polynomial characterizations for some other cases.

cs.DM↗

arXiv2015

On Konig-Egervary Collections of Maximum Critical Independent Sets

Let G be a simple graph with vertex set V(G). A set S is independent if no two vertices from S are adjacent. The graph G is known to be a Konig-Egervary if alpha(G)+mu(G)= |V(G)|, where alpha(G) denotes the size of a maximum independent set and mu(G) is the cardinality of a maximum matching. The number d(X)= |X|-|N(X)| is the difference of X, and an independent set A is critical if d(A) = max{d(I):I is an independent set in G} (Zhang; 1990). Let Omega(G) denote the family of all maximum independent sets. Let us say that a family Gamma of independent sets is a Konig-Egervary collection if |Union of Gamma| + |Intersection of Gamma| = 2alpha(G) (Jarden, Levit, Mandrescu; 2015). In this paper, we show that if the family of all maximum critical independent sets of a graph G is a Konig-Egervary collection, then G is a Konig-Egervary graph. It generalizes one of our conjectures recently validated in (Short; 2015).

cs.DM↗

arXiv2015

Monotonic Properties of Collections of Maximum Independent Sets of a Graph

Let G be a simple graph with vertex set V(G). A subset S of V(G) is independent if no two vertices from S are adjacent. The graph G is known to be a Konig-Egervary if alpha(G) + mu(G)= |V(G)|, where alpha(G) denotes the size of a maximum independent set and mu(G) is the cardinality of a maximum matching. Let Omega(G) denote the family of all maximum independent sets, and f be the function from the set of subcollections Gamma of Omega(G) such that f(Gamma) = (the cardinality of the union of elements of Gamma) + (the cardinality of the intersection of elements of Gamma). Our main finding claims that f is "<<"-increasing, where the preorder {Gamma1} << {Gamma2} means that the union of all elements of {Gamma1} is a subset of the union of all elements of {Gamma2}, while the intersection of all elements of {Gamma2} is a subset of the intersection of all elements of {Gamma1}. Let us say that a family {Gamma} is a Konig-Egervary collection if f(Gamma) = 2*alpha(G). We conclude with the observation that for every graph G each subcollection of a Konig-Egervary collection is Konig-Egervary as well.

cs.DM↗

arXiv2015

Critical and Maximum Independent Sets of a Graph

Let G be a simple graph with vertex set V(G). A subset S of V(G) is independent if no two vertices from S are adjacent. By Ind(G) we mean the family of all independent sets of G while core(G) and corona(G) denote the intersection and the union of all maximum independent sets, respectively. The number d(X)= |X|-|N(X)| is the difference of the set of vertices X, and an independent set A is critical if d(A)=max{d(I):I belongs to Ind(G)} (Zhang, 1990). Let ker(G) and diadem(G) be the intersection and union, respectively, of all critical independent sets of G (Levit and Mandrescu, 2012). In this paper, we present various connections between critical unions and intersections of maximum independent sets of a graph. These relations give birth to new characterizations of Koenig-Egervary graphs, some of them involving ker(G), core(G), corona(G), and diadem(G).

cs.DM↗