Searcharxiv⌕ Search

arXiv subjects

Katarzyna Paluch

Publications and source records attributed to Katarzyna Paluch.

15 recordsLinked to original sources
arXiv2026

New Approximation Algorithms for Maximum Asymmetric Traveling Salesman and Shortest Superstring

In the maximum asymmetric traveling salesman problem (Max ATSP) we are given a complete directed graph with nonnegative weights on the edges and we wish to compute a traveling salesman tour of maximum weight. In this paper we give a fast combinatorial $\frac{7}{10}$-approximation algorithm for Max ATSP. It is based on techniques of {\em eliminating} and {\em diluting} problematic subgraphs with the aid of {\it half-edges} and a method of edge coloring. (A {\it half-edge} of edge $(u,v)$ is informally speaking "either a head or a tail of $(u,v)$".) A novel technique of {\em diluting} a problematic subgraph $S$ consists in a seeming reduction of its weight, which allows its better handling. The current best approximation algorithms for Max ATSP, achieving the approximation guarantee of $\frac 23$, are due to Kaplan, Lewenstein, Shafrir, Sviridenko (2003) and Elbassioni, Paluch, van Zuylen (2012). Using a result by Mucha, which states that an $α$-approximation algorithm for Max ATSP implies a $(2+\frac{11(1-α)}{9-2α})$-approximation algorithm for the shortest superstring problem (SSP), we obtain also a $(2 \frac{33}{76} \approx 2,434)$-approximation algorithm for SSP, beating the previously best known (having an approximation factor equal to $2 \frac{11}{23} \approx 2,4782$.)

cs.DS↗
arXiv2025

Triangle-free 2-matchings

We consider the problem of finding a maximum size triangle-free $2$-matching in a graph $G=(V,E)$. A (simple) $2$-matching is any subset of the edges such that each vertex is incident to at most two edges from the subset. The first polynomial time algorithm for this problem was given by Hartvigsen in 1984 in his PhD thesis and its improved version has been recently published in a journal. We present a different, significantly simpler algorithm with a relatively short proof of correctness. Our algorithm with running time $O(|V||E|)$ is additionally faster than the one by Hartvigsen having running time $O(|V|^3|E|^2)$. It has been proven before that for any triangle-free $2$-matching $M$ which is not maximum the graph contains an $M$-augmenting path, whose application to $M$ results in a bigger triangle-free $2$-matching. A new observation is that the search for an augmenting path $P$ can be restricted to so-called {\em amenable} paths that go through any triangle $t$ contained in $P\cup M$ a limited number of times. Amenable paths can be characterised with the aid of {\em half-edges}. A {\em half-edge} of edge $e$ is, informally speaking, a half of $e$ containing exactly one of its endpoints. Each half-edge serves also as a {\em hinge} - a connector between one pair of edges on an alternating path. To find an amenable augmenting path we thus dynamically remove and re-add half-edges to forbid or allow some edges to be followed by certain others. The existence of amenable augmenting paths follows from our decomposition theorem for triangle-free $2$-matchings. This decomposition theorem is largely the same as the decomposition from versions 1-6 of this paper and is moreover simpler and stronger than the one given by Kobayashi and Noguchi.

cs.DS↗
arXiv2024

Rectangle Tiling Binary Arrays

The problem of rectangle tiling binary arrays is defined as follows. Given an $n \times n$ array $A$ of zeros and ones and a natural number $p$, our task is to partition $A$ into at most $p$ rectangular tiles, so that the maximal weight of a tile is minimized. A tile is any rectangular subarray of $A$. The weight of a tile is the sum of elements that fall within it. We present a linear $(O(n^2))$ time $(\frac{3}{2}+\frac{p^2}{w(A)})$-approximation algorithm (where $\frac{p^2}{w(A)} < \frac{1}{2}$) for this problem, where $w(A)$ denotes the weight of the whole array $A$. This improves on the previously known approximation with the ratio $2$. The result is best possible in the following sense. The algorithm employs the lower bound of $L=\lceil \frac{w(A)}{p} \rceil$, which is the only known and used bound on the optimum in all algorithms for rectangle tiling. We prove that a better approximation factor for the binary \RTILE cannot be achieved using $L$, because there exist arrays, whose every partition contains a tile with weight at least $(\frac{3}{2}+\frac{p^2}{w(A)})L$. We also consider the dual problem of rectangle tiling for binary arrays, where we are given an upper bound on the weight of the tiles, and we have to cover the array $A$ with the minimum number of non-overlapping tiles. Both problems have natural extensions to $d$-dimensional versions, for which we provide analogous results.

cs.CG↗
arXiv2024

Clique-free t-matchings in degree-bounded graphs

We consider problems of finding a maximum size/weight $t$-matching without forbidden subgraphs in an undirected graph $G$ with the maximum degree bounded by $t+1$, where $t$ is an integer greater than $2$. Depending on the variant forbidden subgraphs denote certain subsets of $t$-regular complete partite subgraphs of $G$. A graph is complete partite if there exists a partition of its vertex set such that every pair of vertices from different sets is connected by an edge and vertices from the same set form an independent set. A clique $K_t$ and a bipartite clique $K_{t,t}$ are examples of complete partite graphs. These problems are natural generalizations of the triangle-free and square-free $2$-matching problems in subcubic graphs. In the weighted setting we assume that the weights of edges of $G$ are vertex-induced on every forbidden subgraph. We present simple and fast combinatorial algorithms for these problems. The presented algorithms are the first ones for the weighted versions, and for the unweighted ones, are faster than those known previously. Our approach relies on the use of gadgets with so-called half-edges. A half-edge of edge $e$ is, informally speaking, a half of $e$ containing exactly one of its endpoints.

cs.DS↗
arXiv2021

Optimal General Matchings

Given a graph $G=(V,E)$ and for each vertex $v \in V$ a subset $B(v)$ of the set $\{0,1,\ldots, d_G(v)\}$ a $B$-matching of $G$ is any set $F \subseteq E$ such that $d_F(v) \in B(v)$ for each vertex $v$. The general matching problem asks the existence of a $B$-matching in a given graph. A set $B(v)$ is said to have a {\em gap of length} $p$ if there exists a number $k \in B(v)$ such that $k+1, \ldots, k+p \notin B(v)$ and $k+p+1 \in B(v)$. Without any restrictions the general matching problem is NP-complete. However, if no set $B(v)$ contains a gap of length greater than $1$, then the problem can be solved in polynomial time and Cornuejols \cite{Cor} presented an algorithm for finding a $B$-matching, if it exists. In this paper we consider a version of the general matching problem, in which we are interested in finding a $B$-matching having a maximum (or minimum) number of edges. We present the first polynomial time algorithm for the maximum weight $B$-matching for the case when no set $B(v)$ contains a gap of length greater than $1$.

cs.DS↗
arXiv2020

A simple combinatorial algorithm for restricted 2-matchings in subcubic graphs -- via half-edges

We consider three variants of the problem of finding a maximum weight restricted $2$-matching in a subcubic graph $G$. (A $2$-matching is any subset of the edges such that each vertex is incident to at most two of its edges.) Depending on the variant a restricted $2$-matching means a $2$-matching that is either triangle-free or square-free or both triangle- and square-free. While there exist polynomial time algorithms for the first two types of $2$-matchings, they are quite complicated or use advanced methodology. For each of the three problems we present a simple reduction to the computation of a maximum weight $b$-matching. The reduction is conducted with the aid of half-edges. A half-edge of edge $e$ is, informally speaking, a half of $e$ containing exactly one of its endpoints. For a subset of triangles of $G$, we replace each edge of such a triangle with two half-edges. Two half-edges of one edge $e$ of weight $w(e)$ may get different weights, not necessarily equal to $\frac{1}{2}w(e)$. In the metric setting when the edge weights satisfy the triangle inequality, this has a geometric interpretation connected to how an incircle partitions the edges of a triangle. Our algorithms are additionally faster than those known before. The running time of each of them is $O(n^2\log{n})$, where $n$ denotes the number of vertices in the graph.

cs.DS↗
arXiv2020

The Dynamics of Rank-Maximal and Popular Matchings

Given a bipartite graph, where the two sets of vertices are applicants and posts and ranks on the edges represent preferences of applicants over posts, a {\em rank-maximal} matching is one in which the maximum number of applicants is matched to their rank one posts and subject to this condition, the maximum number of applicants is matched to their rank two posts, and so on. We study the dynamic version of the problem in which a new applicant or post may be added to the graph and we would like to maintain a rank-maximal matching. We show that after the arrival of one vertex, we are always able to update the existing rank-maximal matching in $\mathcal{O}(\min(c'n ,n^2) + m)$ time, where $n$ denotes the number of applicants, $m$ the number of edges and $c'$ the maximum rank of an edge in an optimal solution. Additionally, we update the matching using a minimal number of changes (replacements). All cases of a deletion of a vertex/edge and an addition of an edge can be reduced to the problem of handling the addition of a vertex. As a by-product, we also get an analogous $\mathcal{O}(m)$ result for the dynamic version of the (one-sided) popular matching problem. Our results are based on the novel use of the properties of the Edmonds-Gallai decomposition. The presented ideas may find applications in other (dynamic) matching problems.

cs.DS↗
arXiv2018

Manipulation Strategies for the Rank Maximal Matching Problem

We consider manipulation strategies for the rank-maximal matching problem. In the rank-maximal matching problem we are given a bipartite graph $G = (A \cup P, E)$ such that $A$ denotes a set of applicants and $P$ a set of posts. Each applicant $a \in A$ has a preference list over the set of his neighbours in $G$, possibly involving ties. Preference lists are represented by ranks on the edges - an edge $(a,p)$ has rank $i$, denoted as $rank(a,p)=i$, if post $p$ belongs to one of $a$'s $i$-th choices. A rank-maximal matching is one in which the maximum number of applicants is matched to their rank one posts and subject to this condition, the maximum number of applicants is matched to their rank two posts, and so on. A rank-maximal matching can be computed in $O(\min(c \sqrt{n},n) m)$ time, where $n$ denotes the number of applicants, $m$ the number of edges and $c$ the maximum rank of an edge in an optimal solution. A central authority matches applicants to posts. It does so using one of the rank-maximal matchings. Since there may be more than one rank- maximal matching of $G$, we assume that the central authority chooses any one of them randomly. Let $a_1$ be a manipulative applicant, who knows the preference lists of all the other applicants and wants to falsify his preference list so that he has a chance of getting better posts than if he were truthful. In the first problem addressed in this paper the manipulative applicant $a_1$ wants to ensure that he is never matched to any post worse than the most preferred among those of rank greater than one and obtainable when he is truthful. In the second problem the manipulator wants to construct such a preference list that the worst post he can become matched to by the central authority is best possible or in other words, $a_1$ wants to minimize the maximal rank of a post he can become matched to.

cs.DS↗
arXiv2016

A $4/5$ - Approximation Algorithm for the Maximum Traveling Salesman Problem

In the maximum traveling salesman problem (Max TSP) we are given a complete undirected graph with nonnegative weights on the edges and we wish to compute a traveling salesman tour of maximum weight. We present a fast combinatorial $\frac 45$ - approximation algorithm for Max TSP. The previous best approximation for this problem was $\frac 79$. The new algorithm is based on a novel technique of eliminating difficult subgraphs via half-edges, a new method of edge coloring and a technique of exchanging edges. A half-edge of edge $e=(u,v)$ is informally speaking "a half of $e$ containing either $u$ or $v$".

cs.DS↗
arXiv2015

Characterisation of Strongly Stable Matchings

An instance of a strongly stable matching problem (SSMP) is an undirected bipartite graph $G=(A \cup B, E)$, with an adjacency list of each vertex being a linearly ordered list of ties, which are subsets of vertices equally good for a given vertex. Ties are disjoint and may contain one vertex. A matching $M$ is a set of vertex-disjoint edges. An edge $(x,y) \in E \setminus M$ is a {\em blocking edge} for $M$ if $x$ is either unmatched or strictly prefers $y$ to its current partner in $M$, and $y$ is either unmatched or strictly prefers $x$ to its current partner in $M$ or is indifferent between them. A matching is {\em strongly stable} if there is no blocking edge with respect to it. We present an algorithm for the generation of all strongly stable matchings, thus solving an open problem already stated in the book by Gusfield and Irving \cite{GI}. It has previously been shown that strongly stable matchings form a distributive lattice and although the number of strongly stable matchings can be exponential in the number of vertices, we show that there exists a partial order with $O(m)$ elements representing all strongly stable matchings, where $m$ denotes the number of edges in the graph. We give two algorithms that construct two such representations: one in $O(nm^2)$ time and the other in $O(nm)$ time, where $n$ denotes the number of vertices in the graph. Note that the construction of the second representation has the same time complexity as that of computing a single strongly stable matching.

cs.DS↗
arXiv2014

Maximum ATSP with Weights Zero and One via Half-Edges

We present a fast combinatorial $3/4$-approximation algorithm for the maximum asymmetric TSP with weights zero and one. The approximation factor of this algorithm matches the currently best one given by Bläser in 2004 and based on linear programming. Our algorithm first computes a maximum size matching and a maximum weight cycle cover without certain cycles of length two but possibly with {\em half-edges} - a half-edge of a given edge $e$ is informally speaking a half of $e$ that contains one of the endpoints of $e$. Then from the computed matching and cycle cover it extracts a set of paths, whose weight is large enough to be able to construct a traveling salesman tour with the claimed guarantee.

cs.DS↗
arXiv2014

Faster and simpler approximation of stable matchings

We give a 3/2-approximation algorithm for stable matchings that runs in $O(m)$ time. The previously best known algorithm by McDermid has the same approximation ratio but runs in $O(n^{3/2}m)$ time, where $n$ denotes the number of people and $m$ is the total length of the preference lists in a given instance. Also the algorithm and the analysis are much simpler. We also give the extension of the algorithm for the many-to-many setting. (This is the version of the paper from March 2011)

cs.DS↗
arXiv2014

Better Approximation Algorithms for Maximum Asymmetric Traveling Salesman and Shortest Superstring

In the maximum asymmetric traveling salesman problem (Max ATSP) we are given a complete directed graph with nonnegative weights on the edges and we wish to compute a traveling salesman tour of maximum weight. In this paper we give a fast combinatorial $\frac 34$-approximation algorithm for Max ATSP. It is based on a novel use of {\it half-edges}, matchings and a new method of edge coloring. (A {\it half-edge} of edge $(u,v)$ is informally speaking "either a head or a tail of $(u,v)$".) The current best approximation algorithms for Max ATSP, achieving the approximation guarantee of $\frac 23$, are due to Kaplan, Lewenstein, Shafrir and Sviridenko and Elbassioni, Paluch, van Zuylen. Using a recent result by Mucha, which states that an $α$-approximation algorithm for Max ATSP implies a $(2+\frac{11(1-α)}{9-2α})$-approximation algorithm for the shortest superstring problem (SSP), we obtain also a $(2 \frac{11}{30} \approx 2,3667)$-approximation algorithm for SSP, beating the previously best known (having approximation factor equal to $2 \frac{11}{23} \approx 2,4782$.)

cs.DS↗
arXiv2010

Popular b-matchings

Suppose that each member of a set of agents has a preference list of a subset of houses, possibly involving ties and each agent and house has their capacity denoting the maximum number of correspondingly agents/houses that can be matched to him/her/it. We want to find a matching $M$, for which there is no other matching $M'$ such that more agents prefer $M'$ to $M$ than $M$ to $M'$. (What it means that an agent prefers one matching to the other is explained in the paper.) Popular matchings have been studied quite extensively, especially in the one-to-one setting. We provide a characterization of popular b-matchings for two defintions of popularity, show some $NP$-hardness results and for certain versions describe polynomial algorithms.

cs.DS↗