In the d-Clique Packing problem, given a graph G and an integer t, we need to decide whether G contains a set of t pairwise vertex-disjoint cliques of size d each. This generalizes Triangle Packing and it is NP-complete for all d ≥ 3. For each such d, we show how to solve the problem in n^𝒪(k^{d-1}) time where k is the clique-width of the graph (with a k-expression of G given in the input). We complement this by showing that, assuming the Exponential-Time Hypothesis (ETH), there is no algorithm that solves the problem in n^o(k^{d-1}) time for any fixed d ≥ 3, already for the special case of seeking a partition into cliques of size d. Our proof also entails W[1]-hardness of d-Clique Packing (and d-Clique Partitioning) parameterized by clique-width for each d ≥ 3. Our work continues a series of results on ETH-tight bounds for fundamental graph problems started by Fomin et al. (SICOMP 2010+2014) who obtained tight bounds for Max-Cut and Edge Dominating Set.

A large number of NP-hard graph problems can be solved in c^w n^𝒪(1) time and space when the input graph is provided together with a tree decomposition of width w. Moreover, assuming the Strong Exponential-Time Hypothesis (SETH) we have essentially matching lower bounds for many such problems. The main drawback of these results is that the corresponding dynamic programming algorithms use exponential space, which makes them infeasible for larger w, and there is some evidence that this cannot be avoided. This motivates using somewhat more restrictive structure/decompositions of the graph to also get good (exponential) dependence on the corresponding parameter but use only polynomial space. A number of papers have contributed to this quest by studying problems relative to treedepth, and have obtained fast polynomial space algorithms, often matching the dependence on treewidth in the time bound. E.g., a number of connectivity problems could be solved by adapting the cut-and-count technique of Cygan et al. (FOCS 2011, TALG 2022) to treedepth, but this excluded well-known path and cycle problems such as Hamiltonian Cycle (Hegerfeld and Kratsch, STACS 2020). Recently, Nederlof et al. (SIDMA 2023) showed how to solve Hamiltonian Cycle, and several related problems, in 5^τ n^𝒪(1) randomized time and polynomial space when provided with an elimination forest of depth τ. We present a faster (also randomized) algorithm, running in 4^τ n^𝒪(1) time and polynomial space, for the same set of problems. We use ordered pairs of what we call consistent matchings, rather than perfect matchings in an auxiliary graph, to get the improved time bound.

We introduce a new notion of acyclicity representation in labeled graphs, and present three applications thereof. Our main result is an algorithm that, given a graph $G$ and a $k$-clique expression of $G$, in time $O(6^kn^c)$ counts modulo $2$ the number of feedback vertex sets of $G$ of each size. We achieve this through an involved subroutine for merging partial solutions at union nodes in the expression. In the usual way this results in a one-sided error Monte-Carlo algorithm for solving the decision problem in the same time. We complement these by a matching lower bound under the Strong Exponential-Time Hypothesis (SETH). This closes an open question that appeared multiple times in the literature [ESA 23, ICALP 24, IPEC 25]. We also present an algorithm that, given a graph $G$ and a tree decomposition of width $k$ of $G$, in time $O(3^kn^c)$ counts modulo $2$ the number of feedback vertex sets of $G$ of each size. This matches the known SETH-tight bound for the decision version, which was obtained using the celebrated cut-and-count technique [FOCS 11, TALG 22]. Unlike other applications of cut-and-count, which use the isolation lemma to reduce a decision problem to counting solutions modulo $2$, this bound was obtained via counting other objects, leaving the complexity of counting solutions modulo $2$ open. Finally, we present a one-sided error Monte-Carlo algorithm that, given a graph $G$ and a $k$-clique expression of $G$, in time $O(18^kn^c)$ decides the existence of a connected feedback vertex set of size $b$ in $G$. We provide a matching lower bound under SETH.

In this work we contribute to the study of the fine-grained complexity of problems parameterized by multi-clique-width, which was initiated by Fürer [ITCS 2017] and pursued further by Chekan and Kratsch [MFCS 2023]. Multi-clique-width is a parameter defined analogously to clique-width but every vertex is allowed to hold multiple labels simultaneously. This parameter is upper-bounded by both clique-width and treewidth (plus a constant), hence it generalizes both of them without an exponential blow-up. Conversely, graphs of multi-clique-width k have clique-width at most 2^k, and there exist graphs with clique-width at least 2^{Ω(k)}. Thus, while the two parameters are functionally equivalent, the fine-grained complexity of problems may differ relative to them. As our first and main result we show that under ETH the Max Cut problem cannot be solved in time n^{2^{o(k)}} ⋅ f(k) on graphs of multi-clique-width k for any computable function f. For clique-width k an n^{𝒪(k)} algorithm by Fomin et al. [SIAM J. Comput. 2014] is tight under ETH. This makes Max Cut the first known problem for which the tight running times differ for parameterization by clique-width and multi-clique-width and it contributes to the short list of known lower bounds of form n^{2^{o(k)}} ⋅ f(k). As our second contribution we show that Hamiltonian Cycle and Edge Dominating Set can be solved in time n^{𝒪(k)} on graphs of multi-clique-width k matching the tight running time for clique-width. These results answer three questions left open by Chekan and Kratsch [MFCS 2023].

A stable cutset in a graph G is a set S ⊆ V ( G ) such that vertices of S are pairwise non-adjacent and such that G − S is disconnected, i.e., it is both stable (or independent) set and a cutset (or separator). Unlike general cutsets, it is NP -complete to determine whether a given graph G has any stable cutset. Recently, Rauch et al. [FCT 2023 & JCSS 2025] gave a number of fixed-parameter tractable (FPT) algorithms, running in time f ( k ) ⋅ | V ( G ) | c , for Stable Cutset under a variety of parameters k such as the size of a (given) dominating set, the size of an odd cycle transversal, or the deletion distance to P 5 -free graphs. Earlier works imply FPT algorithms relative to clique-width and relative to solution size. We complement these findings by giving the first results on the existence of polynomial kernelizations for Stable Cutset , i.e., efficient preprocessing algorithms that return an equivalent instance of size polynomial in the parameter value. Under the standard assumption that NP ⊈ coNP/poly , we show that no polynomial kernelization is possible relative to the deletion distance to a single path, generalizing deletion distance to various graph classes, nor by the size of a (given) dominating set. We also show that under the same assumption no polynomial kernelization is possible relative to solution size, i.e., given ( G , k ) answering whether there is a stable cutset of size at most k . On the positive side, we show polynomial kernelizations for parameterization by modulators to a single clique, to a cluster or a co-cluster graph, and by twin cover.

We extend the notion of lossy kernelization, introduced by Lokshtanov et al. (2017) [19] , to approximate Turing kernelization. An α -approximate Turing kernelization for a parameterized optimization problem is a polynomial-time algorithm that, when given access to an oracle that outputs c -approximate solutions in O ( 1 ) time, computes an α ⋅ c -approximate solution to the considered problem, using calls to the oracle of size at most f ( k ) for some function f that only depends on the parameter. Using this definition, we show that Independent Set parameterized by treewidth ℓ has a ( 1 + ε ) -approximate Turing kernelization with O ( ℓ 2 ε ) vertices, answering an open question posed by Lokshtanov et al. (2017) [19] . Furthermore, we give ( 1 + ε ) -approximate Turing kernelizations for the following graph problems parameterized by treewidth: Vertex Cover , Edge Clique Cover , Edge-Disjoint Triangle Packing , and Connected Vertex Cover . We generalize the result for Independent Set and Vertex Cover by showing that all graph problems that we will call friendly admit ( 1 + ε ) -approximate Turing kernelizations of polynomial size when parameterized by treewidth. We use this to establish approximate Turing kernelizations for Vertex-Disjoint H -packing for connected graphs H , Clique Cover , Feedback Vertex Set , and Edge Dominating Set .

Cutwidth is a widely studied parameter and it quantifies how well a graph can be decomposed along small edge-cuts. It complements pathwidth, which captures decomposition by small vertex separators, and it is well-known that cutwidth upper-bounds pathwidth. The SETH-tight parameterized complexity of problems on graphs of bounded pathwidth (and treewidth) has been actively studied over the past decade while for cutwidth the complexity of many classical problems remained open. For Hamiltonian Cycle, it is known that a (2+√2)^{pw} n^𝒪(1) algorithm is optimal for pathwidth under SETH [Cygan et al. JACM 2018]. Van Geffen et al. [J. Graph Algorithms Appl. 2020] and Bojikian et al. [STACS 2023] asked which running time is optimal for this problem parameterized by cutwidth. We answer this question with (1+√2)^{ctw} n^𝒪(1) by providing matching upper and lower bounds. Second, as our main technical contribution, we close the gap left by van Heck [2018] for Partition Into Triangles (and Triangle Packing) by improving both upper and lower bound and getting a tight bound of ∛{3}^{ctw} n^𝒪(1), which to our knowledge exhibits the only known tight non-integral basis apart from Hamiltonian Cycle [Cygan et al. JACM 2018] and C₄-Hitting Set [SODA 2025]. We show that the cuts inducing a disjoint union of paths of length three (unions of so-called Z-cuts) lie at the core of the complexity of the problem - usually lower-bound constructions use simpler cuts inducing either a matching or a disjoint union of bicliques. Finally, we determine the optimal running times for Max Cut (2^{ctw} n^𝒪(1)) and Induced Matching (3^{ctw} n^𝒪(1)) by providing matching lower bounds for the existing algorithms - the latter result also answers an open question for treewidth by Chaudhary and Zehavi [WG 2023].

A kernelization is an efficient algorithm that given an instance of a parameterized problem returns an equivalent instance of size bounded by some function of the input parameter value. It is quite well understood which problems do or (conditionally) do not admit a kernelization where this size bound is polynomial, a so-called polynomial kernelization. Unfortunately, such polynomial kernelizations are known only in fairly restrictive settings where a small parameter value corresponds to a strong restriction on the global structure on the instance. Motivated by this, Antipov and Kratsch [WG 2025] proposed a local variant of kernelization, called boundaried kernelization, that requires only local structure to achieve a local improvement of the instance, which is in the spirit of protrusion replacement used in meta-kernelization [Bodlaender et al.\ JACM 2016]. They obtain polynomial boundaried kernelizations as well as (unconditional) lower bounds for several well-studied problems in kernelization. In this work, we leverage the matroid-based techniques of Kratsch and Wahlström [JACM 2020] to obtain randomized polynomial boundaried kernelizations for \smultiwaycut, \dtmultiwaycut, \oddcycletransversal, and \vertexcoveroct, for which randomized polynomial kernelizations in the usual sense were known before. A priori, these techniques rely on the global connectivity of the graph to identify reducible (irrelevant) vertices. Nevertheless, the separation of the local part by its boundary turns out to be sufficient for a local application of these methods.

Recently, Bojikian and Kratsch [2023] have presented a novel approach to tackle connectivity problems parameterized by clique-width ($\operatorname{cw}$), based on counting small representations of partial solutions (modulo two). Using this technique, they were able to get a tight bound for the Steiner Tree problem, answering an open question posed by Hegerfeld and Kratsch [ESA, 2023]. We use the same technique to solve the Connected Odd Cycle Transversal problem in time $\mathcal{O}^*(12^{\operatorname{cw}})$. We define a new representation of partial solutions by separating the connectivity requirement from the 2-colorability requirement of this problem. Moreover, we prove that our result is tight by providing SETH-based lower bound excluding algorithms with running time $\mathcal{O}^*((12-ε)^{\operatorname{lcw}})$ even when parameterized by linear clique-width. This answers the second question posed by Hegerfeld and Kratsch in the same paper.

We show a flow-augmentation algorithm in directed graphs: There exists a randomized polynomial-time algorithm that, given a directed graph G , two vertices s, t ∈ V(G) , and an integer k , adds (randomly) to G a number of arcs such that for every minimal st -cut Z in G of size at most k , with probability 2 −poly( k ) the set Z becomes a minimum st -cut in the resulting graph. We also provide a deterministic counterpart of this procedure. The directed flow-augmentation tool allows us to prove fixed-parameter tractability of a number of problems parameterized by the cardinality of the deletion set whose parameterized complexity status was repeatedly posed as open problems: Chain SAT , defined by Chitnis, Egri, and Marx [ESA’13, Algorithmica’17], a number of weighted variants of classic directed cut problems, such as Weighted st - Cut or Weighted Directed Feedback Vertex Set . By proving that Chain SAT is FPT, we confirm a conjecture of Chitnis, Egri, and Marx that, for any graph H , if the List H - Coloring problem is polynomial-time solvable, then the corresponding vertex-deletion problem is fixed-parameter tractable.

Abstract. We study the parameterized problem of satisfying “almost all” constraints of a given formula [Formula: see text] over a fixed, finite Boolean constraint language [Formula: see text], with or without weights. More precisely, for each finite Boolean constraint language [Formula: see text], we consider the following two problems. In Min SAT([Formula: see text]), the input is a formula [Formula: see text] over [Formula: see text] and an integer [Formula: see text], and the task is to find an assignment [Formula: see text] that satisfies all but at most [Formula: see text] constraints of [Formula: see text], or determine that no such assignment exists. In Weighted Min SAT([Formula: see text]), the input additionally contains a weight function [Formula: see text] and an integer [Formula: see text], and the task is to find an assignment [Formula: see text] such that (1) [Formula: see text] satisfies all but at most [Formula: see text] constraints of [Formula: see text], and (2) the total weight of the violated constraints is at most [Formula: see text]. We give a complete dichotomy for the fixed-parameter tractability of these problems: We show that for every Boolean constraint language [Formula: see text], either Weighted Min SAT([Formula: see text]) is [Formula: see text]; or Weighted Min SAT([Formula: see text]) is [Formula: see text]-hard but Min SAT([Formula: see text]) is [Formula: see text]; or Min SAT([Formula: see text]) is [Formula: see text]-hard. This generalizes recent work of Kim et al. [in SODA 2021, SIAM, Philadelphia, 2021, pp. 149–168], which did not consider weighted problems and only considered languages [Formula: see text] that cannot express implications [Formula: see text] (as is used to, e.g., model digraph cut problems). Our result generalizes and subsumes multiple previous results, including the FPT algorithms for Weighted Almost 2-SAT, weighted and unweighted [Formula: see text]-Chain SAT, and Coupled Min-Cut, as well as weighted and directed versions of the latter. The main tool used in our algorithms is the recently developed method of directed flow-augmentation [E. J. Kim et al., in STOC 2022, ACM, 2022, pp. 938–947].

Given a graph G = (V,E), a set T ⊆ V, and an integer b, the Steiner Tree problem asks whether G has a connected subgraph H with at most b vertices that spans all of T. This work presents a 3^k⋅ n^𝒪(1) time one-sided Monte-Carlo algorithm for solving Steiner Tree when additionally a clique-expression of width k is provided. Known lower bounds for less expressive parameters imply that this dependence on the clique-width of G is optimal assuming the Strong Exponential-Time Hypothesis (SETH). Indeed our work establishes that the parameter dependence of Steiner Tree is the same for any graph parameter between cutwidth and clique-width, assuming SETH. Our work contributes to the program of determining the exact parameterized complexity of fundamental hard problems relative to structural graph parameters such as treewidth, which was initiated by Lokshtanov et al. [SODA 2011 & TALG 2018] and which by now has seen a plethora of results. Since the cut-and-count framework of Cygan et al. [FOCS 2011 & TALG 2022], connectivity problems have played a key role in this program as they pose many challenges for developing tight upper and lower bounds. Recently, Hegerfeld and Kratsch [ESA 2023] gave the first application of the cut-and-count technique to problems parameterized by clique-width and obtained tight bounds for Connected Dominating Set and Connected Vertex Cover, leaving open the complexity of other benchmark connectivity problems such as Steiner Tree and Feedback Vertex Set. Our algorithm for Steiner Tree does not follow the cut-and-count technique and instead works with the connectivity patterns of partial solutions. As a first technical contribution we identify a special family of so-called complete patterns that has strong (existential) representation properties, and using these at least one solution will be preserved. Furthermore, there is a family of 3^k basis patterns that (parity) represents the complete patterns, i.e., it has the same number of solutions modulo two. Our main technical contribution, a new technique called "isolating a representative," allows us to leverage both forms of representation (existential and parity). Both complete patterns and isolation of a representative will likely be applicable to other (connectivity) problems.

We present an undirected version of the recently introduced flow-augmentation technique: Given an undirected multigraph G with distinguished vertices s,t ∈ V(G) and an integer k , one can in randomized k 𝒪(1) ⋅ (|V(G)| + |E(G)|) time sample a set A ⊆ \(\binom{V(G)}{2}\) such that the following holds: for every inclusion-wise minimal st -cut Z in G of cardinality at most k , Z becomes a minimum-cardinality cut between s and t in G+A (i.e., in the multigraph G with all edges of A added) with probability 2 -𝒪( k log k ). Compared to the version for directed graphs [STOC 2022], the version presented here has improved success probability (2 -𝒪( k log k ) instead of 2 -𝒪( k 4 log k ) ), linear dependency on the graph size in the running time bound, and an arguably simpler proof. An immediate corollary is that the Bi-objective st -Cut problem can be solved in randomized FPT time 2 𝒪( k log k ) (|V(G)|+|E(G)|) on undirected graphs.

The complexity of problems involving global constraints is usually much more difficult to understand than the complexity of problems only involving local constraints. A natural form of global constraints are connectivity constraints. We study connectivity problems from a fine-grained parameterized perspective. In a breakthrough, Cygan et al. (TALG 2022) first obtained algorithms with single-exponential running time c^{tw} n^O(1) for connectivity problems parameterized by treewidth by introducing the cut-and-count-technique. Furthermore, the obtained bases c were shown to be optimal under the Strong Exponential-Time Hypothesis (SETH). However, since only sparse graphs may admit small treewidth, we lack knowledge of the fine-grained complexity of connectivity problems with respect to dense structure. The most popular graph parameter to measure dense structure is arguably clique-width, which intuitively measures how easily a graph can be constructed by repeatedly adding bicliques. Bergougnoux and Kanté (TCS 2019) have shown, using the rank-based approach, that also parameterized by clique-width many connectivity problems admit single-exponential algorithms. Unfortunately, the obtained running times are far from optimal under SETH. We show how to obtain optimal running times parameterized by clique-width for two benchmark connectivity problems, namely Connected Vertex Cover and Connected Dominating Set. These are the first tight results for connectivity problems with respect to clique-width and these results are obtained by developing new algorithms based on the cut-and-count-technique and novel lower bound constructions. Precisely, we show that there exist one-sided error Monte-Carlo algorithms that given a k-clique-expression solve Connected Vertex Cover in time 6^k n^O(1), and Connected Dominating Set in time 5^k n^O(1). Both results are shown to be tight under SETH.

In this work, we study two natural generalizations of clique-width introduced by Martin Fürer. Multi-clique-width (mcw) allows every vertex to hold multiple labels [ITCS 2017], while for fusion-width (fw) we have a possibility to merge all vertices of a certain label [LATIN 2014]. Fürer has shown that both parameters are upper-bounded by treewidth thus making them more appealing from an algorithmic perspective than clique-width and asked for applications of these parameters for problem solving. First, we determine the relation between these two parameters by showing that $\operatorname{mcw} \leq \operatorname{fw} + 1$. Then we show that when parameterized by multi-clique-width, many problems (e.g., Connected Dominating Set) admit algorithms with the same running time as for clique-width despite the exponential gap between these two parameters. For some problems (e.g., Hamiltonian Cycle) we show an analogous result for fusion-width: For this we present an alternative view on fusion-width by introducing so-called glue-expressions which might be interesting on their own. All algorithms obtained in this work are tight up to (Strong) Exponential Time Hypothesis.

We study the parameterized problem of satisfying “almost all” constraints of a given formula F over a fixed, finite Boolean constraint language Γ, with or without weights. More precisely, for each finite Boolean constraint language Γ, we consider the following two problems. In MIN SAT(T), the input is a formula F over Γ and an integer k, and the task is to find an assignment α : V(F) → {0,1} that satisfies all but at most k constraints of F, or determine that no such assignment exists. In WEIGHTED MIN SAT(Γ), the input additionally contains a weight function ω : F → ℤ+ and an integer W, and the task is to find an assignment α such that (1) α satisfies all but at most k constraints of F, and (2) the total weight of the violated constraints is at most W. We give a complete dichotomy for the fixed-parameter tractability of these problems: We show that for every Boolean constraint language Γ, either WEIGHTED MIN SAT(Γ) is FPT; or WEIGHTED MIN SAT(Γ) is W[1]-hard but MIN SAT(Γ) is FPT; or MIN SAT (Γ) is W[1]-hard. This generalizes recent work of Kim et al. (SODA 2021) which did not consider weighted problems, and only considered languages Γ that cannot express implications (u → v) (as is used to, e.g., model digraph cut problems). Our result generalizes and subsumes multiple previous results, including the FPT algorithms for WEIGHTED Almost 2-SAT, weighted and unweighted ℓ-CHAIN SAT, and COUPLED MIN-CUT, as well as weighted and directed versions of the latter. The main tool used in our algorithms is the recently developed method of directed flow-augmentation (Kim et al., STOC 2022). * This research is a part of a project that have received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme Grant Agreement 714704 (M. Pilipczuk). Eun Jung Kim is supported by the grant from French National Research Agency under JCJC program (ASSK: ANR-18-CE40-0025-01). The full version of the paper can be accessed at https://arxiv.org/abs/2207.07422.

In this work we start the investigation of tight complexity bounds for connectivity problems parameterized by cutwidth assuming the Strong Exponential-Time Hypothesis (SETH). Van Geffen et al. posed this question for odd cycle transversal and feedback vertex set. We answer it for these two and four further problems, namely connected vertex cover, connected domintaing set, steiner tree, and connected odd cycle transversal. For the latter two problems it sufficed to prove lower bounds that match the running time inherited from parameterization by treewidth; for the others we provide faster algorithms than relative to treewidth and prove matching lower bounds. For upper bounds we first extend the idea of Groenland et al.~[STACS~2022] to solve what we call coloring-like problem. Such problems are defined by a symmetric matrix $M$ over $\mathbb{F}_2$ indexed by a set of colors. The goal is to count the number (modulo some prime $p$) of colorings of a graph such that $M$ has a $1$-entry if indexed by the colors of the end-points of any edge. We show that this problem can be solved faster if $M$ has small rank over $\mathbb{F}_p$. We apply this result to get our upper bounds for connected vertex cover and connected dominating set. The upper bounds for odd cycle transversal and feedback vertex set use a subdivision trick to get below the bounds that matrix rank would yield.

Parameterized complexity seeks to use input structure to obtain faster algorithms for NP-hard problems. This has been most successful for graphs of low treewidth: Many problems admit fast algorithms relative to treewidth and many of them are optimal under SETH. Fewer such results are known for more general structure such as low clique-width and more restrictive structure such as low deletion distance to a sparse graph class. Despite these successes, such results remain "islands'' within the realm of possible structure. Rather than adding more islands, we seek to determine the transitions between them, that is, we aim for structural thresholds where the complexity increases as input structure becomes more general. Going from deletion distance to treewidth, is a single deletion set to a graph with simple components enough to yield the same lower bound as for treewidth or does it take many disjoint separators? Going from treewidth to clique-width, how much more density entails the same complexity as clique-width? Conversely, what is the most restrictive structure that yields the same lower bound? For treewidth, we obtain both refined and new lower bounds that apply already to graphs with a single separator $X$ such that $G-X$ has treewidth $r=O(1)$, while $G$ has treewidth $|X|+O(1)$. We rule out algorithms running in time $O^*((r+1-ε)^{k})$ for Deletion to $r$-Colorable parameterized by $k=|X|$. For clique-width, we rule out time $O^*((2^r-ε)^k)$ for Deletion to $r$-Colorable, where $X$ is now allowed to consist of $k$ twinclasses. There are further results on Vertex Cover, Dominating Set and Maximum Cut. All lower bounds are matched by existing and newly designed algorithms.

We show a flow-augmentation algorithm in directed graphs: There exists a randomized polynomial-time algorithm that, given a directed graph G, two integers s,t ∈ V(G), and an integer k, adds (randomly) to G a number of arcs such that for every minimal st-cut Z in G of size at most k, with probability 2−poly(k) the set Z becomes a minimum st-cut in the resulting graph.

The Vertex Cover problem plays an essential role in the study of polynomial kernelization in parameterized complexity, i.e., the study of provable and efficient preprocessing for ${\mathsf{NP}}$-hard problems. Motivated by the great variety of positive and negative results for kernelization for Vertex Cover subject to different parameters and graph classes, we seek to unify and generalize them using so-called blocking sets. A blocking set is a set of vertices such that no optimal vertex cover contains all vertices in the blocking set, and the study of minimal blocking sets played implicit and explicit roles in many existing results. We show that in the most-studied setting, parameterized by the size of a deletion set to a specified graph class ${\mathcal{C}}$, bounded minimal blocking set size is necessary but not sufficient to get a polynomial kernelization. Under mild technical assumptions, bounded minimal blocking set size is shown to allow an essentially tight polynomial-time reduction in the number of connected components. We then determine the exact maximum size of minimal blocking sets for graphs of bounded elimination distance to any hereditary class $\mathcal{C}$, including the case of graphs of bounded treedepth. We get similar but not tight bounds for certain nonhereditary classes $\mathcal{C}$, including the class ${\mathcal{C}}_{{\mathrm{LP}}}$ of graphs where integral and fractional vertex cover size coincide. These bounds allow us to derive polynomial kernels for Vertex Cover parameterized by the size of a deletion set to graphs of bounded elimination distance to, e.g., forest, bipartite, or ${\mathcal{C}}_{\mathrm{LP}}$ graphs.

We present a new technique for designing fixed-parameter algorithms for graph cut problems in undirected graphs, which we call flow augmentation. Our technique is applicable to problems that can be phrased as a search for an (edge) (s, t)-cut of cardinality at most k in an undirected graph G with designated terminals s and t. More precisely, we consider problems where an (unknown) solution is a set Z ⊆ E(G) of size at most k such that •in G – Z, s and t are in distinct connected components,•every edge of Z connects two distinct connected components of G – Z, and•if we define the set Zs, t ⊆ Z as those edges e ∊ Z for which there exists an (s, t)-path Pe with E(Pe) ∩ Z = {e}, then Zs, t separates s from t. We prove that in the above scenario one can in randomized time k(1) (|V(G)| + |E(G)|) add a number of edges to the graph so that with probably at least 2–(k log k) no added edge connects two components of G – Z, and Zs, t becomes a minimum cut between s and t. This additional property becomes a handy lever in applications. For example, consider the question of an (s, t)-cut of cardinality at most k and of minimum possible weight (assuming edge weights in G). While the problem is NP-hard in general, it easily reduces to the maximum flow / minimum cut problem if we additionally assume that k is the minimum possible cardinality of an (s, t)-cut in G. Hence, we immediately obtain that the aforementioned problem admits an 2(k log k) n(1)-time randomized fixed-parameter algorithm. We apply our method to obtain a randomized fixed-parameter algorithm for a notorious “hard nut” graph cut problem we call Coupled Min-Cut. This problem emerges out of the study of FPT algorithms for Min CSP problems (see below), and was unamenable to other techniques for parameterized algorithms in graph cut problems, such as Randomized Contractions, Treewidth Reduction or Shadow Removal. In fact, we go one step further. To demonstrate the power of the approach, we consider more generally the Boolean Min CSP(Γ)-problems, a.k.a. Min SAT(Γ), parameterized by the solution cost. This is a framework of optimization problems that includes problems such as Almost 2-SAT and the notorious i-Chain SAT problem. We are able to show that every problem Min SAT(Γ) is either (1) FPT, (2) W[1]-hard, or (3) able to express the soft constraint (u → v), and thereby also the min-cut problem in directed graphs. All the W[1]-hard cases were known or immediate, and the main new result is an FPT algorithm for a generalization of Coupled Min-Cut. In other words, flow-augmentation is powerful enough to let us solve every fixed-parameter tractable problem in the class, except those that explicitly encompass directed graph cuts.

Given two disjoint sets W1 and W2 of points in the plane, the Optimal Discretization problem asks for the minimum size of a family of horizontal and vertical lines that separate W1 from W2, that is, in every region into which the lines partition the plane there are either only points of W1, or only points of W2, or the region is empty. Equivalently, Optimal Discretization can be phrased as a task of discretizing continuous variables: We would like to discretize the range of x-coordinates and the range of y-coordinates into as few segments as possible, maintaining that no pair of points from W1 × W2 are projected onto the same pair of segments under this discretization. We provide a fixed-parameter algorithm for the problem, parameterized by the number of lines in the solution. Our algorithm works in time , where k is the bound on the number of lines to find and n is the number of points in the input. Our result answers in positive a question of Bonnet, Giannopolous, and Lampis [IPEC 2017] and of Froese (PhD thesis, 2018) and is in contrast with the known intractability of two closely related generalizations: the Rectangle Stabbing problem and the generalization in which the selected lines are not required to be axis-parallel.

We extend the notion of lossy kernelization, introduced by Lokshtanov et al. [STOC 2017], to approximate Turing kernelization. An $α$-approximate Turing kernel for a parameterized optimization problem is a polynomial-time algorithm that, when given access to an oracle that outputs $c$-approximate solutions in $O(1)$ time, obtains an $(α\cdot c)$-approximate solution to the considered problem, using calls to the oracle of size at most $f(k)$ for some function $f$ that only depends on the parameter. Using this definition, we show that Independent Set parameterized by treewidth $\ell$ has a $(1+\varepsilon)$-approximate Turing kernel with $O(\frac{\ell^2}{\varepsilon})$ vertices, answering an open question posed by Lokshtanov et al. [STOC 2017]. Furthermore, we give $(1+\varepsilon)$-approximate Turing kernels for the following graph problems parameterized by treewidth: Vertex Cover, Edge Clique Cover, Edge-Disjoint Triangle Packing and Connected Vertex Cover. We generalize the result for Independent Set and Vertex Cover, by showing that all graph problems that we will call "friendly" admit $(1+\varepsilon)$-approximate Turing kernels of polynomial size when parameterized by treewidth. We use this to obtain approximate Turing kernels for Vertex-Disjoint $H$-packing for connected graphs $H$, Clique Cover, Feedback Vertex Set and Edge Dominating Set.

A breakthrough result of Cygan et al. (FOCS 2011) showed that connectivity problems parameterized by treewidth can be solved much faster than the previously best known time ?^*(2^{?(twlog tw)}). Using their inspired Cut&Count technique, they obtained ?^*(α^tw) time algorithms for many such problems. Moreover, they proved these running times to be optimal assuming the Strong Exponential-Time Hypothesis. Unfortunately, like other dynamic programming algorithms on tree decompositions, these algorithms also require exponential space, and this is widely believed to be unavoidable. In contrast, for the slightly larger parameter called treedepth, there are already several examples of matching the time bounds obtained for treewidth, but using only polynomial space. Nevertheless, this has remained open for connectivity problems. In the present work, we close this knowledge gap by applying the Cut&Count technique to graphs of small treedepth. While the general idea is unchanged, we have to design novel procedures for counting consistently cut solution candidates using only polynomial space. Concretely, we obtain time ?^*(3^d) and polynomial space for Connected Vertex Cover, Feedback Vertex Set, and Steiner Tree on graphs of treedepth d. Similarly, we obtain time ?^*(4^d) and polynomial space for Connected Dominating Set and Connected Odd Cycle Transversal.

The Vertex Cover problem plays an essential role in the study of polynomial kernelization in parameterized complexity, i.e., the study of provable and efficient preprocessing for NP-hard problems. Motivated by the great variety of positive and negative results for kernelization for Vertex Cover subject to different parameters and graph classes, we seek to unify and generalize them using so-called blocking sets. A blocking set is a set of vertices such that no optimal vertex cover contains all vertices in the blocking set, and the study of minimal blocking sets played implicit and explicit roles in many existing results. We show that in the most-studied setting, parameterized by the size of a deletion set to a specified graph class ?, bounded minimal blocking set size is necessary but not sufficient to get a polynomial kernelization. Under mild technical assumptions, bounded minimal blocking set size is showed to allow an essentially tight efficient reduction in the number of connected components. We then determine the exact maximum size of minimal blocking sets for graphs of bounded elimination distance to any hereditary class ?, including the case of graphs of bounded treedepth. We get similar but not tight bounds for certain non-hereditary classes ?, including the class ?_{LP} of graphs where integral and fractional vertex cover size coincide. These bounds allow us to derive polynomial kernels for Vertex Cover parameterized by the size of a deletion set to graphs of bounded elimination distance to, e.g., forest, bipartite, or ?_{LP} graphs.

Computing all-pairs shortest paths is a fundamental and much-studied problem with many applications. Unfortunately, despite intense study, there are still no significantly faster algorithms for it than the $\mathcal{O}(n^3)$ time algorithm due to Floyd and Warshall (1962). Somewhat faster algorithms exist for the vertex-weighted version if fast matrix multiplication may be used. Yuster (SODA 2009) gave an algorithm running in time $\mathcal{O}(n^{2.842})$, but no combinatorial, truly subcubic algorithm is known. Motivated by the recent framework of efficient parameterized algorithms (or "FPT in P"), we investigate the influence of the graph parameters clique-width ($cw$) and modular-width ($mw$) on the running times of algorithms for solving All-Pairs Shortest Paths. We obtain efficient (and combinatorial) parameterized algorithms on non-negative vertex-weighted graphs of times $\mathcal{O}(cw^2n^2)$, resp. $\mathcal{O}(mw^2n + n^2)$. If fast matrix multiplication is allowed then the latter can be improved to $\mathcal{O}(mw^{1.842}n + n^2)$ using the algorithm of Yuster as a black box. The algorithm relative to modular-width is adaptive, meaning that the running time matches the best unparameterized algorithm for parameter value $mw$ equal to $n$, and they outperform them already for $mw \in \mathcal{O}(n^{1 - \varepsilon})$ for any $\varepsilon > 0$.

In this paper, we study multi-budgeted variants of the classic minimum cut problem and graph separation problems that turned out to be important in parameterized complexity: Skew Multicut and Directed Feedback Arc Set. In our generalization, we assign colors $$1,2,\ldots ,\ell $$ to some edges and give separate budgets $$k_{1},k_{2},\ldots ,k_{\ell }$$ for colors $$1,2,\ldots ,\ell $$ . For every color $$i\in \{1,\ldots ,\ell \}$$ , let $$E_{i}$$ be the set of edges of color i. The solution C for the multi-budgeted variant of a graph separation problem not only needs to satisfy the usual separation requirements (i.e., be a cut, a skew multicut, or a directed feedback arc set, respectively), but also needs to satisfy that $$|C\cap E_{i}|\le k_{i}$$ for every $$i\in \{1,\ldots ,\ell \}$$ . Contrary to the classic minimum cut problem, the multi-budgeted variant turns out to be NP-hard even for $$\ell = 2$$ . We propose FPT algorithms parameterized by $$k=k_{1}+\cdots +k_{\ell }$$ for all three problems. To this end, we develop a branching procedure for the multi-budgeted minimum cut problem that measures the progress of the algorithm not by reducing k as usual, by but elevating the capacity of some edges and thus increasing the size of maximum source-to-sink flow. Using the fact that a similar strategy is used to enumerate all important separators of a given size, we merge this process with the flow-guided branching and show an FPT bound on the number of (appropriately defined) important multi-budgeted separators. This allows us to extend our algorithm to the Skew Multicut and Directed Feedback Arc Set problems. Furthermore, we show connections of the multi-budgeted variants with weighted variants of the directed cut problems and the Chain $$\ell $$ -SAT problem, whose parameterized complexity remains an open problem. We show that these problems admit a bounded-in-parameter number of “maximally pushed” solutions (in a similar spirit as important separators are maximally pushed), giving somewhat weak evidence towards their tractability.

We continue the development of matroid-based techniques for kernelization, initiated by the present authors [47]. We significantly extend the usefulness of matroid theory in kernelization by showing applications of a result on representative sets due to Lovász [51] and Marx [53]. As a first result, we show how representative sets can be used to derive a polynomial kernel for the elusive ALMOST 2- SAT problem (where the task is to remove at most k clauses to make a 2- CNF formula satisfiable), solving a major open problem in kernelization. This result also yields a new O(√log OPT)-approximation for the problem, improving on the O(√log n)-approximation of Agarwal et al. [3] and an implicit O(log OPT)-approximation due to Even et al. [24]. We further apply the representative sets tool to the problem of finding irrelevant vertices in graph cut problems, that is, vertices that can be made undeletable without affecting the answer to the problem. This gives the first significant progress towards a polynomial kernel for the MULTIWAY CUT problem; in particular, we get a kernel of O( k s+1 ) vertices for MULTIWAY CUT instances with at most s terminals. Both these kernelization results have significant spin-off effects, producing the first polynomial kernels for a range of related problems. More generally, the irrelevant vertex results have implications for covering min cuts in graphs. For a directed graph G=(V,E) and sets S, T ⊆ V , let r be the size of a minimum ( S,T )-vertex cut (which may intersect S and T ). We can find a set Z ⊆ V of size O(|S| . |T| . r) that contains a minimum ( A,B )-vertex cut for every A ⊆ S , B ⊆ T . Similarly, for an undirected graph G=(V,E) , a set of terminals X ⊆ V , and a constant s , we can find a set Z ⊆ V of size O(|X| s+1 ) that contains a minimum multiway cut for every partition of X into at most s pairwise disjoint subsets. Both results are polynomial time. We expect this to have further applications; in particular, we get direct, reduction rule-based kernelizations for all problems above, in contrast to the indirect compression-based kernel previously given for ODD CYCLE TRANSVERSAL [47]. All our results are randomized, with failure probabilities that can be made exponentially small in n , due to needing a representation of a matroid to apply the representative sets tool.

The area of parameterized approximation seeks to combine approximation and parameterized algorithms to obtain, e.g., (1+epsilon)-approximations in f(k,epsilon)n^O(1) time where k is some parameter of the input. The goal is to overcome lower bounds from either of the areas. We obtain the following results on parameterized approximability: - In the maximum independent set of rectangles problem (MISR) we are given a collection of n axis parallel rectangles in the plane. Our goal is to select a maximum-cardinality subset of pairwise non-overlapping rectangles. This problem is NP-hard and also W[1]-hard [Marx, ESA'05]. The best-known polynomial-time approximation factor is O(log log n) [Chalermsook and Chuzhoy, SODA'09] and it admits a QPTAS [Adamaszek and Wiese, FOCS'13; Chuzhoy and Ene, FOCS'16]. Here we present a parameterized approximation scheme (PAS) for MISR, i.e. an algorithm that, for any given constant epsilon>0 and integer k>0, in time f(k,epsilon)n^g(epsilon), either outputs a solution of size at least k/(1+epsilon), or declares that the optimum solution has size less than k. - In the (2-dimensional) geometric knapsack problem (2DK) we are given an axis-aligned square knapsack and a collection of axis-aligned rectangles in the plane (items). Our goal is to translate a maximum cardinality subset of items into the knapsack so that the selected items do not overlap. In the version of 2DK with rotations (2DKR), we are allowed to rotate items by 90 degrees. Both variants are NP-hard, and the best-known polynomial-time approximation factor is 2+epsilon [Jansen and Zhang, SODA'04]. These problems admit a QPTAS for polynomially bounded item sizes [Adamaszek and Wiese, SODA'15]. We show that both variants are W[1]-hard. Furthermore, we present a PAS for 2DKR. For all considered problems, getting time f(k,epsilon)n^O(1), rather than f(k,epsilon)n^g(epsilon), would give FPT time f'(k)n^O(1) exact algorithms by setting epsilon=1/(k+1), contradicting W[1]-hardness. Instead, for each fixed epsilon>0, our PASs give (1+epsilon)-approximate solutions in FPT time. For both MISR and 2DKR our techniques also give rise to preprocessing algorithms that take n^g(epsilon) time and return a subset of at most k^g(epsilon) rectangles/items that contains a solution of size at least k/(1+epsilon) if a solution of size k exists. This is a special case of the recently introduced notion of a polynomial-size approximate kernelization scheme [Lokshtanov et al., STOC'17].

In the fundamental Maximum Matching problem the task is to find a maximum cardinality set of pairwise disjoint edges in a given undirected graph. The fastest algorithm for this problem, due to Micali and Vazirani, runs in time $\mathcal{O}(\sqrt{n}m)$ and stands unbeaten since 1980. It is complemented by faster, often linear-time, algorithms for various special graph classes. Moreover, there are fast parameterized algorithms, e.g., time $\mathcal{O}(km\log n)$ relative to tree-width $k$, which outperform $\mathcal{O}(\sqrt{n}m)$ when the parameter is sufficiently small. We show that the Micali-Vazirani algorithm, and in fact any algorithm following the phase framework of Hopcroft and Karp, is adaptive to beneficial input structure. We exhibit several graph classes for which such algorithms run in linear time $\mathcal{O}(n+m)$. More strongly, we show that they run in time $\mathcal{O}(\sqrt{k}m)$ for graphs that are $k$ vertex deletions away from any of several such classes, without explicitly computing an optimal or approximate deletion set; before, most such bounds were at least $Ω(km)$. Thus, any phase-based matching algorithm with linear-time phases obliviously interpolates between linear time for $k=\mathcal{O}(1)$ and the worst case of $\mathcal{O}(\sqrt{n}m)$ when $k=Θ(n)$. We complement our findings by proving that the phase framework by itself still allows $Ω(\sqrt{n})$ phases, and hence time $Ω(\sqrt{n}m)$, even on paths, cographs, and bipartite chain graphs.

In the NP-hard Edge Dominating Set problem (EDS) we are given a graph $G=(V,E)$ and an integer $k$, and need to determine whether there is a set $F\subseteq E$ of at most $k$ edges that are incident with all (other) edges of $G$. It is known that this problem is fixed-parameter tractable and admits a polynomial kernel when parameterized by $k$. A caveat for this parameter is that it needs to be large, i.e., at least equal to half the size of a maximum matching of $G$, for instances not to be trivially negative. Motivated by this, we study the existence of polynomial kernels for EDS when parameterized by structural parameters that may be much smaller than $k$. Unfortunately, at first glance this looks rather hopeless: Even when parameterized by the deletion distance to a disjoint union of paths $P_3$ of length two there is no polynomial kernelization (under standard assumptions), ruling out polynomial kernels for many smaller parameters like the feedback vertex set size. In contrast, somewhat surprisingly, there is a polynomial kernelization for deletion distance to a disjoint union of paths $P_5$ of length four. As our main result, we fully classify for all finite sets $\mathcal{H}$ of graphs, whether a kernel size polynomial in $|X|$ is possible when given $X$ such that each connected component of $G-X$ is isomorphic to a graph in $\mathcal{H}$.

We consider the problem of finding a subcomplex $\mathcal{K}'$ of a simplicial complex $\mathcal{K}$ such that $\mathcal{K}'$ is homeomorphic to the 2-dimensional sphere, $\mathbb{S}^2$. We study two variants of this problem. The first asks if there exists such a $\mathcal{K}'$ with at most $\mathcal{K}$ triangles, and we show that this variant is ${\mathsf{W[1]}}$-hard and, assuming the exponential time hypothesis, admits no $n^{o(\sqrt{k})}$-time algorithm. We also give an algorithm that is tight with regard to this lower bound. The second problem is the dual of the first and asks if $\mathcal{K}'$ can be found by removing at most $k$ triangles from $\mathcal{K}$. This variant has an immediate $\mathcal{O}(3^{k}poly(|\mathcal{K}|))$-time algorithm, and we show that it admits a polynomial kernelization to $\mathcal{O}(k^2)$ triangles, as well as a polynomial compression to a weighted version with bit-size $\mathcal{O}(k \log k)$. This article has been changed.

We study the influence of a graph parameter called modular-width on the time complexity for optimally solving well-known polynomial problems such as Maximum Matching, Triangle Counting, and Maximum $s$-$t$ Vertex-Capacitated Flow. The modular-width of a graph depends on its (unique) modular decomposition tree, and can be computed in linear time $O(n+m)$ for graphs with $n$ vertices and $m$ edges. Modular decompositions are an important tool for graph algorithms, e.g., for linear-time recognition of certain graph classes. Throughout, we obtain efficient parameterized algorithms of running times $O(f(mw)n+m)$, $O(n+f(mw)m)$ , or $O(f(mw)+n+m)$ for graphs of modular-width $mw$. Our algorithm for Maximum Matching, running in time $O(mw^2\log mw \cdot n+m)$, is both faster and simpler than the recent $O(mw^4n+m)$ time algorithm of Coudert et al. (SODA 2018). For several other problems, e.g., Triangle Counting and Maximum $b$-Matching, we give adaptive algorithms, meaning that their running times match the best unparameterized algorithms for worst-case modular-width of $mw=Θ(n)$ and they outperform them already for $mw=o(n)$, until reaching linear time for $mw=O(1)$.

In this paper, we study multi-budgeted variants of the classic minimum cut problem and graph separation problems that turned out to be important in parameterized complexity: Skew Multicut and Directed Feedback Arc Set. In our generalization, we assign colors 1,2,...,l to some edges and give separate budgets k_1,k_2,...,k_l for colors 1,2,...,l. For every color i in {1,...,l}, let E_i be the set of edges of color i. The solution C for the multi-budgeted variant of a graph separation problem not only needs to satisfy the usual separation requirements (i.e., be a cut, a skew multicut, or a directed feedback arc set, respectively), but also needs to satisfy that |C cap E_i| <= k_i for every i in {1,...,l}. Contrary to the classic minimum cut problem, the multi-budgeted variant turns out to be NP-hard even for l = 2. We propose FPT algorithms parameterized by k=k_1 +...+ k_l for all three problems. To this end, we develop a branching procedure for the multi-budgeted minimum cut problem that measures the progress of the algorithm not by reducing k as usual, by but elevating the capacity of some edges and thus increasing the size of maximum source-to-sink flow. Using the fact that a similar strategy is used to enumerate all important separators of a given size, we merge this process with the flow-guided branching and show an FPT bound on the number of (appropriately defined) important multi-budgeted separators. This allows us to extend our algorithm to the Skew Multicut and Directed Feedback Arc Set problems. Furthermore, we show connections of the multi-budgeted variants with weighted variants of the directed cut problems and the Chain l-SAT problem, whose parameterized complexity remains an open problem. We show that these problems admit a bounded-in-parameter number of "maximally pushed" solutions (in a similar spirit as important separators are maximally pushed), giving somewhat weak evidence towards their tractability.

For an even integer t ≥ 2, the Matching Connectivity matrix H t is a matrix that has rows and columns both labeled by all perfect matchings of the complete graph on t vertices; an entry H t [ M 1 , M 2 ] is 1 if M 1 and M 2 form a Hamiltonian cycle and 0 otherwise. Motivated by applications for the Hamiltonicity problem, we show that H t has rank exactly 2 t /2−1 over GF(2). The upper bound is established by an explicit factorization of H t as the product of two submatrices; the matchings labeling columns and rows, respectively, of the submatrices therefore form a basis X t of H t . The lower bound follows because the 2 t /2−1 × 2 t /2−1 submatrix with rows and columns labeled by X t can be seen to have full rank. We obtain several algorithmic results based on the rank of H t and the particular structure of the matchings in X t . First, we present a 1.888 n n O (1) time Monte Carlo algorithm that solves the Hamiltonicity problem in directed bipartite graphs. Second, we give a Monte Carlo algorithm that solves the problem in (2 + √ 2) pw n O (1) time when provided with a path decomposition of width pw for the input graph. Moreover, we show that this algorithm is best possible under the Strong Exponential Time Hypothesis, in the sense that an algorithm with running time (2 + √2 − ϵ) pw n O (1) , for any ϵ > 0, would imply the breakthrough result of a (2 − ϵ ′ ) n -time algorithm for CNF-Sat for some ϵ ′ > 0.

The subset feedback vertex set problem generalizes the classical feedback vertex set problem and asks, for a given undirected graph G = (V, E), a set S ⊆ V, and an integer k, whether there exists a set X of at most k vertices such that no cycle in G − X contains a vertex of S. It was independently shown by Cygan et al. (ICALP ’11, SIDMA ’13) and Kawarabayashi and Kobayashi (JCTB ’12) that subset feedback vertex set is fixed-parameter tractable for parameter k. Cygan et al. asked whether the problem also admits a polynomial kernelization. We answer the question of Cygan et al. positively by giving a randomized polynomial kernelization for the equivalent version where S is a set of edges. In a first step we show that edge subset feedback vertex set has a randomized polynomial kernel parameterized by |S| + k with $\mathcal {O}(|S|^{2}k)$ vertices. For this we use the matroid-based tools of Kratsch and Wahlström (FOCS ’12) that for example were used to obtain a polynomial kernel for s-multiway cut. Next we present a preprocessing that reduces the given instance (G, S, k) to an equivalent instance (G′, S′, k′) where the size of S′ is bounded by $\mathcal {O}(k^{4})$ . These two results lead to a randomized polynomial kernel for subset feedback vertex set with $\mathcal {O}(k^{9})$ vertices.

In the vertex cover problem we are given a graph $G=(V,E)$ and an integer $k$ and have to determine whether there is a set $X\subseteq V$ of size at most $k$ such that each edge in $E$ has at least one endpoint in $X$. The problem can be easily solved in time $\mathcal{O}^*(2^k)$, making it fixed-parameter tractable (FPT) with respect to $k$. While the fastest known algorithm takes only time $\mathcal{O}^*(1.2738^k)$, much stronger improvements have been obtained by studying parameters that are smaller than $k$. Apart from treewidth-related results, the arguably best algorithm for vertex cover runs in time $\mathcal{O}^*(2.3146^p)$, where $p=k-LP(G)$ is only the excess of the solution size $k$ over the best fractional vertex cover [Lokshtanov et al. ACM Trans. Algorithms, 11 (2014), pp. 15:1--15:31]. Since $p\leq k$ but $k$ cannot be bounded in terms of $p$ alone, this strictly increases the range of tractable instances. Recently, [Garg and Philip, Proceedings of the Symposium on Discrete Algorithms, 2016, pp. 1152--1166] greatly contributed to understanding the parameterized complexity of the vertex cover problem. They prove that $2LP(G)-MM(G)$ is a lower bound for the vertex cover size of $G$, where $MM(G)$ is the size of a largest matching of $G$, and proceed to study parameter $\ell=k-(2LP(G)-MM(G))$. They give an algorithm of running time $\mathcal{O}^*(3^\ell)$, proving that vertex cover is FPT in $\ell$. It can be easily observed that $\ell\leq p$ whereas $p$ cannot be bounded in terms of $\ell$ alone. We complement the work of Garg and Philip by proving that vertex cover admits a randomized polynomial kernelization in terms of $\ell$, i.e., an efficient preprocessing to size polynomial in $\ell$. This improves over parameter $p=k-LP(G)$ for which this was previously known [Kratsch and Wahlström, Proceedings of the IEEE Symposium on Foundations of Computer Science, 2012, pp. 450--459].

In a Stackelberg Pricing Game a distinguished player, the leader, chooses prices for a set of items, and the other players, the followers, each seeks to buy a minimum cost feasible subset of the items. The goal of the leader is to maximize her revenue, which is determined by the sold items and their prices. Most previously studied cases of such games can be captured by a combinatorial model where we have a base set of items, some with fixed prices, some priceable, and constraints on the subsets that are feasible for each follower. In this combinatorial setting, Briest et al. and Balcan et al. independently showed that the maximum revenue can be approximated to a factor of H_k ~ log(k), where k is the number of priceable items. Our results are twofold. First, we strongly generalize the model by letting the follower minimize any continuous function plus a linear term over any compact subset of R_(n>=0); the coefficients (or prices) in the linear term are chosen by the leader and determine her revenue. In particular, this includes the fundamental case of linear programs. We give a tight lower bound on the revenue of the leader, generalizing the results of Briest et al. and Balcan et al. Besides, we prove that it is strongly NP-hard to decide whether the optimum revenue exceeds the lower bound by an arbitrarily small factor. Second, we study the parameterized complexity of computing the optimal revenue with respect to the number k of priceable items. In the combinatorial setting, given an efficient algorithm for optimal follower solutions, the maximum revenue can be found by enumerating the 2^k subsets of priceable items and computing optimal prices via a result of Briest et al., giving time O(2^k|I|^c ) where |I| is the input size. Our main result here is a W[1]-hardness proof for the case where the followers minimize a linear program, ruling out running time f(k)|I|^c unless FPT = W[1] and ruling out time |I|^o(k) under the Exponential-Time Hypothesis.

We revisit the topic of polynomial kernels for Vertex Cover relative to structural parameters. Our starting point is a recent paper due to Fomin and Strømme [WG 2016] who gave a kernel with O(|X|^{12}) vertices when X is a vertex set such that each connected component of G-X contains at most one cycle, i.e., X is a modulator to a pseudoforest. We strongly generalize this result by using modulators to d-quasi-forests, i.e., graphs where each connected component has a feedback vertex set of size at most d, and obtain kernels with O(|X|^{3d+9}) vertices. Our result relies on proving that minimal blocking sets in a d-quasi-forest have size at most d+2. This bound is tight and there is a related lower bound of O(|X|^{d+2-epsilon}) on the bit size of kernels. In fact, we also get bounds for minimal blocking sets of more general graph classes: For d-quasi-bipartite graphs, where each connected component can be made bipartite by deleting at most d vertices, we get the same tight bound of d+2 vertices. For graphs whose connected components each have a vertex cover of cost at most d more than the best fractional vertex cover, which we call d-quasi-integral, we show that minimal blocking sets have size at most 2d+2, which is also tight. Combined with existing randomized polynomial kernelizations this leads to randomized polynomial kernelizations for modulators to d-quasi-bipartite and d-quasi-integral graphs. There are lower bounds of O(|X|^{d+2-epsilon}) and O(|X|^{2d+2-epsilon}) for the bit size of such kernels.

We consider the problem of finding a subcomplex K' of a simplicial complex K such that K' is homeomorphic to the 2-dimensional sphere, S^2. We study two variants of this problem. The first asks if there exists such a K' with at most k triangles, and we show that this variant is W[1]-hard and, assuming ETH, admits no O(n^{o(sqrt(k))}) time algorithm. We also give an algorithm that is tight with regards to this lower bound. The second problem is the dual of the first, and asks if K' can be found by removing at most k triangles from K. This variant has an immediate O(3^k poly(|K|)) time algorithm, and we show that it admits a polynomial kernelization to O(k^2) triangles, as well as a polynomial compression to a weighted version with bit-size O(k log k).

Binary search finds a given element in a sorted array with an optimal number of log n queries. However, binary search fails even when the array is only slightly disordered or access to its elements is subject to errors. We study the worst-case query complexity of search algorithms that are robust to imprecise queries and that adapt to perturbations of the order of the elements. We give (almost) tight results for various parameters that quantify query errors and that measure array disorder. In particular, we exhibit settings where query complexities of log n + ck, (1+epsilon) log n + ck, and sqrt(cnk)+o(nk) are best-possible for parameter value k, any epsilon > 0, and constant c.

In the Vertex Cover problem we are given a graph G=(V,E) and an integer k and have to determine whether there is a set X subseteq V of size at most k such that each edge in E has at least one endpoint in X. The problem can be easily solved in time O^*(2^k), making it fixed-parameter tractable (FPT) with respect to k. While the fastest known algorithm takes only time O^*(1.2738^k), much stronger improvements have been obtained by studying parameters that are smaller than k. Apart from treewidth-related results, the arguably best algorithm for Vertex Cover runs in time O^*(2.3146^p), where p = k - LP(G) is only the excess of the solution size k over the best fractional vertex cover [Lokshtanov et al., TALG 2014]. Since p <= k but k cannot be bounded in terms of p alone, this strictly increases the range of tractable instances. Recently, Garg and Philip (SODA 2016) greatly contributed to understanding the parameterized complexity of the Vertex Cover problem. They prove that 2LP(G) - MM(G) is a lower bound for the vertex cover size of G, where MM(G) is the size of a largest matching of G, and proceed to study parameter l = k - (2LP(G)-MM(G)). They give an algorithm of running time O^*(3^l), proving that Vertex Cover is FPT in l. It can be easily observed that l <= p whereas p cannot be bounded in terms of l alone. We complement the work of Garg and Philip by proving that Vertex Cover admits a randomized polynomial kernelization in terms of l, i.e., an efficient preprocessing to size polynomial in l. This improves over parameter p = k - LP(G) for which this was previously known [Kratsch and Wahlström, FOCS 2012].

We continue the study of preprocessing under uncertainty that was initiated independently by Assadi et al. (FSTTCS 2015) and Fafianie et al. (STACS 2016). Here, we are given an instance of a tractable problem with a large static/known part and a small part that is dynamic/uncertain, and ask if there is an efficient algorithm that computes an instance of size polynomial in the uncertain part of the input, from which we can extract an optimal solution to the original instance for all (usually exponentially many) instantiations of the uncertain part. In the present work, we focus on the Matroid Intersection problem. Amongst others we present a positive preprocessing result for the important case of finding a largest common independent set in two linear matroids. Motivated by an application for intersecting two gammoids we also revisit Maximum Flow. There we tighten a lower bound of Assadi et al. and give an alternative positive result for the case of low uncertain capacity that yields a Maximum Flow instance as output rather than a matrix.

In this paper, we introduce a new model for sublinear algorithms called \emph{dynamic sketching}. In this model, the underlying data is partitioned into a large \emph{static} part and a small \emph{dynamic} part and the goal is to compute a summary of the static part (i.e, a \emph{sketch}) such that given any \emph{update} for the dynamic part, one can combine it with the sketch to compute a given function. We say that a sketch is \emph{compact} if its size is bounded by a polynomial function of the length of the dynamic data, (essentially) independent of the size of the static part. A graph optimization problem $P$ in this model is defined as follows. The input is a graph $G(V,E)$ and a set $T \subseteq V$ of $k$ terminals; the edges between the terminals are the dynamic part and the other edges in $G$ are the static part. The goal is to summarize the graph $G$ into a compact sketch (of size poly$(k)$) such that given any set $Q$ of edges between the terminals, one can answer the problem $P$ for the graph obtained by inserting all edges in $Q$ to $G$, using only the sketch. We study the fundamental problem of computing a maximum matching and prove tight bounds on the sketch size. In particular, we show that there exists a (compact) dynamic sketch of size $O(k^2)$ for the matching problem and any such sketch has to be of size $Ω(k^2)$. Our sketch for matchings can be further used to derive compact dynamic sketches for other fundamental graph problems involving cuts and connectivities. Interestingly, our sketch for matchings can also be used to give an elementary construction of a \emph{cut-preserving vertex sparsifier} with space $O(kC^2)$ for $k$-terminal graphs; here $C$ is the total capacity of the edges incident on the terminals. Additionally, we give an improved lower bound (in terms of $C$) of $Ω(C/\log{C})$ on size of cut-preserving vertex sparsifiers.

The input to the NP-hard point line cover problem (PLC) consists of a set P of n points on the plane and a positive integer k ; the question is whether there exists a set of at most k lines that pass through all points in P . By straightforward reduction rules, one can efficiently reduce any input to one with at most k 2 points. We show that this easy reduction is already essentially tight under standard assumptions. More precisely, unless the polynomial hierarchy collapses to its third level, for any ϵ > 0, there is no polynomial-time algorithm that reduces every instance ( P , k ) of PLC to an equivalent instance with O ( k 2 −ϵ) points. This answers, in the negative, an open problem posed by Lokshtanov [2009]. Our proof uses the notion of a kernel from parameterized complexity, and the machinery for deriving lower bounds on the size of kernels developed by Dell and van Melkebeek [2010, 2014]. It has two main ingredients: We first show, by reduction from vertex cover , that—unless the polynomial hierarchy collapses—PLC has no kernel of total size O ( k 2 −ϵ) bits. This does not directly imply the claimed lower bound on the number of points , since the best-known polynomial-time encoding of a PLC instance with n points requires ω( n 2 ) bits. To get around this hurdle, we build on work of Alon [1986] and devise an oracle communication protocol of cost O ( n log n ) for PLC. This protocol, together with the lower bound on the total size (which also holds for such protocols), yields the stated lower bound on the number of points. While a number of essentially tight polynomial lower bounds on total sizes of kernels are known, our result is—to the best of our knowledge—the first to show a nontrivial lower bound for structural/secondary parameters. It is also the first example of a lower bound for kernelization that makes use of the full power of the oracle communication protocol lower bounds that can be obtained from the work of Dell and van Melkebeek. We combine the main abstract ideas of our proof to derive a general recipe that could be used to obtain such lower bounds for other problems with unknown or insufficiently strong encodings.

We study the NP-complete Minimum Shared Edges (MSE) problem. Given an undirected graph, a source and a sink vertex, and two integers p and k, the question is whether there are p paths in the graph connecting the source with the sink and sharing at most k edges. Herein, an edge is shared if it appears in at least two paths. We show that MSE is W[1]-hard when parameterized by the treewidth of the input graph and the number k of shared edges combined. We show that MSE is fixed-parameter tractable with respect to p, but does not admit a polynomial-size kernel (unless NP is a subset of coNP/poly). In the proof of the fixed-parameter tractability of MSE parameterized by p, we employ the treewidth reduction technique due to Marx, O'Sullivan, and Razgon [ACM TALG 2013].

In the Vector Connectivity problem we are given an undirected graph G=(V,E), a demand function phi: V => {0,...,d}, and an integer k. The question is whether there exists a set S of at most k vertices such that every vertex v in V\S has at least phi(v) vertex-disjoint paths to S; this abstractly captures questions about placing servers in a network, or warehouses on a map, relative to demands. The problem is NP-hard already for instances with d=4 (Cicalese et al., Theor. Comput. Sci. 2015), admits a log-factor approximation (Boros et al., Networks 2014), and is fixed-parameter tractable in terms of k (Lokshtanov, unpublished 2014). We prove several results regarding kernelization and approximation for Vector Connectivity and the variant Vector d-Connectivity where the upper bound d on demands is a constant. For Vector d-Connectivity we give a factor d-approximation algorithm and construct a vertex-linear kernelization, i.e., an efficient reduction to an equivalent instance with f(d)k=O(k) vertices. For Vector Connectivity we get a factor opt-approximation and we show that it has no kernelization to size polynomial in k+d unless NP \subseteq coNP/poly, making f(d)\poly(k) optimal for Vector d-Connectivity. Finally, we provide a write-up for fixed-parameter tractability of Vector Connectivity(k) by giving a different algorithm based on matroid intersection.

The Multicut problem, given a graph G, a set of terminal pairs $\mathcal{T}=\{(s_i,t_i)\ |\ 1\leq i\leq r\}$, and an integer $p$, asks whether one can find a cutset consisting of at most $p$ nonterminal vertices that separates all the terminal pairs, i.e., after removing the cutset, $t_i$ is not reachable from $s_i$ for each $1\leq i\leq r$. The fixed-parameter tractability of Multicut in undirected graphs, parameterized by the size of the cutset only, has been recently proved by Marx and Razgon [SIAM J. Comput., 43 (2014), pp. 355--388] and, independently, by Bousquet, Daligault, and Thomassé [Proceedings of STOC, ACM, 2011, pp. 459--468], after resisting attacks as a long-standing open problem. In this paper we prove that Multicut is fixed-parameter tractable on directed acyclic graphs when parameterized both by the size of the cutset and the number of terminal pairs. We complement this result by showing that this is implausible for parameterization by the size of the cutset only, as this version of the problem remains $W[1]$-hard.

The input to the NP-hard Point Line Cover problem (PLC) consists of a set of n points on the plane and a positive integer k, and the question is whether there exists a set of at most k lines which pass through all points in . By straightforward reduction rules one can efficiently reduce any input to one with at most k2 points. We show that this easy reduction is already essentially tight under standard assumptions. More precisely, unless the polynomial hierarchy collapses to its third level, for any ∊ > 0, there is no polynomial-time algorithm that reduces every instance ( , k) of PLC to an equivalent instance with (k2–∊) points. This answers, in the negative, an open problem posed by Lokshtanov (PhD Thesis, 2009). Our proof uses the notion of a kernel from parameterized complexity, and the machinery for deriving lower bounds on the size of kernels developed by Dell and van Melkebeek (STOC 2010). It has two main ingredients: We first show, by reduction from Vertex Cover, that—unless the polynomial hierarchy collapses—PLC has no kernel of total size (k2–∊) bits. This does not directly imply the claimed lower bound on the number of points, since the best known polynomial-time encoding of a PLC instance with n points requires ω(n2) bits. To get around this hurdle we build on work of Goodman, Pollack and Sturmfels (STOC 1989) and devise an oracle communication protocol of cost (nlogn) for PLC; its main building blocks are a bound of (nO(n)) for the order types of n points that are not necessarily in general position and an explicit (albeit slow) algorithm that enumerates a superset of size nO(n) of all possible order types of n points. This protocol, together with the lower bound on the total size (which also holds for such protocols), yields the stated lower bound on the number of points. While a number of essentially tight polynomial lower bounds on total sizes of kernels are known, our result is—to the best of our knowledge—the first to show a nontrivial lower bound for structural/secondary parameters.

Assume that each of n voters may or may not approve each of m issues. If an agent (the lobby) may influence up to k voters, then the central question of the NP-hard Lobbying problem is whether the lobby can choose the voters to be influenced so that as a result each issue gets a majority of approvals. This problem can be modeled as a simple matrix modification problem: Can one replace k rows of a binary n x m-matrix by k all-1 rows such that each column in the resulting matrix has a majority of 1s? Significantly extending on previous work that showed parameterized intractability (W[2]-completeness) with respect to the number k of modified rows, we study how natural parameters such as n, m, k, or the "maximum number of 1s missing for any column to have a majority of 1s" (referred to as "gap value g") govern the computational complexity of Lobbying. Among other results, we prove that Lobbying is fixed-parameter tractable for parameter m and provide a greedy logarithmic-factor approximation algorithm which solves Lobbying even optimally if m < 5. We also show empirically that this greedy algorithm performs well on general instances. As a further key result, we prove that Lobbying is LOGSNP-complete for constant values g>0, thus providing a first natural complete problem from voting for this complexity class of limited nondeterminism.

We introduce the framework of cross-composition for proving kernelization lower bounds. A classical problem $L$ \and/or-cross-composes into a parameterized problem $\mathcal{Q}$ if it is possible to efficiently construct an instance of $\mathcal{Q}$ with polynomially bounded parameter value that expresses the logical and or or of a sequence of instances of $L$. Building on work by Bodlaender et al. and using results of Fortnow and Santhanam, Dell and van Melkebeek, and Drucker, we show that if an NP-hard problem and/or-cross-composes into a parameterized problem $\mathcal{Q}$, then $\mathcal{Q}$ does not admit a polynomial kernel unless $\mbox{NP}\subseteq \mbox{coNP/poly}$ and the polynomial hierarchy collapses. Our technique generalizes and strengthens the techniques of using composition algorithms and of transferring the lower bounds via polynomial parameter transformations. We show its applicability by proving kernelization lower bounds for a number of important graphs problems with structural (nonstandard) parameterizations, e.g., Clique, Chromatic Number, Weighted Feedback Vertex Set, and Weighted Odd Cycle Transversal do not admit polynomial kernels with respect to the vertex cover number of the input graphs unless the polynomial hierarchy collapses, contrasting the fact that these problems are trivially fixed-parameter tractable for this parameter. We have similar lower bounds for Feedback Vertex Set and Odd Cycle Transversal under structural parameterizations. After learning of our results, several teams of authors have successfully applied the cross-composition framework to different parameterized problems. For completeness, our presentation of the framework includes several extensions based on this follow-up work. For example, we show how a relaxed version of or-cross-compositions may be used to give lower bounds on the degree of the polynomial in the kernel size.

The field of kernelization offers a rigorous way of studying the ubiquitous technique of data reduction and preprocessing for combinatorially hard problems. A widely accepted definition of useful data reduction is that of a polynomial kernelization where the output instance is guaranteed to be of size polynomial in some parameter of the input. The fairly recent development of a framework for kernelization lower bounds has made this notion even more attractive as we can now classify many problems into admitting or not admitting polynomial kernelizations. The central notion of the framework is that of a polynomial-time composition algorithm due to Bodlaender et al. (ICALP 2008, JSCC 2009): given t input instances, an or -composition algorithm returns a single-output instance with bounded parameter value that is yes if and only if one of t input instances is yes; it encodes the logical OR of the input instances. Based on a result of Fortnow and Santhanam (STOC 2008, JSCC 2011), Bodlaender et al. show that an or -composition for an NP-hard problem rules out polynomial kernelizations for it unless NP ⊆ coNP/poly (which is known to imply a collapse of the polynomial hierarchy). It is implicit in the work of Fortnow and Santhanam that even co-nondeterministic composition algorithms suffice to rule out polynomial kernelizations. This was first observed in unpublished work of Chen and Müller, and it is an explicit conclusion of recent results by Dell and van Melkebeek (STOC 2010). However, in contrast to the numerous applications of deterministic composition, the added power of co-nondeterminism has not yet been harnessed to obtain kernelization lower bounds. In this work, we present the first example of how co-nondeterminism can help to make a composition algorithm. We study the existence of polynomial kernels for a Ramsey-type problem where, given a graph G and an integer k , the question is whether G contains an independent set or a clique of size at least k . It was asked by Rod Downey whether this problem admits a polynomial kernelization with respect to k ; such a result would greatly speed up the computation of Ramsey numbers. We provide a co-nondeterministic composition based on embedding t instances into a single host graph H . The crux is that the host graph H needs to observe a bound of ℓ ∈ O (log t ) on both its maximum independent set and maximum clique size, while also having a cover of its vertex set by independent sets and cliques all of size ℓ; the co-nondeterministic composition is built around the search for such graphs. Thus, we show that, unless NP ⊆ coNP/poly, the problem does not admit a kernelization with polynomial size guarantee.

The Odd Cycle Transversal problem (OCT) asks whether a given undirected graph can be made bipartite by deleting at most k of its vertices. In a breakthrough result, Reed, Smith, and Vetta (Operations Research Letters, 2004) gave a O (4 k kmn) time algorithm for it; this also implies that instances of the problem can be reduced to a so-called problem kernel of size O (4 k ). Since then, the existence of a polynomial kernel for OCT (i.e., a kernelization with size bounded polynomially in k ) has turned into one of the main open questions in the study of kernelization, open even for the special case of planar input graphs. This work provides the first (randomized) polynomial kernelization for OCT. We introduce a novel kernelization approach based on matroid theory, where we encode all relevant information about a problem instance into a matroid with a representation of size polynomial in k . This represents the first application of matroid theory to kernelization.

In the Correlation Clustering problem, also known as Cluster Editing, we are given an undirected graph G and a positive integer k; the task is to decide whether G can be transformed into a cluster graph, i.e., a disjoint union of cliques, by changing at most k adjacencies, that is, by adding or deleting at most k edges. The motivation of the problem stems from various tasks in computational biology (Ben-Dor et al., Journal of Computational Biology 1999) and machine learning (Bansal et al., Machine Learning 2004). Although in general Correlation Clustering is APX-hard (Charikar et al., FOCS 2003), the version of the problem where the number of cliques may not exceed a prescribed constant p admits a PTAS (Giotis and Guruswami, SODA 2006). We study the parameterized complexity of Correlation Clustering with this restriction on the number of cliques to be created. We give an algorithm that - in time O(2^{O(sqrt{pk})} + n+m) decides whether a graph G on n vertices and m edges can be transformed into a cluster graph with exactly p cliques by changing at most k adjacencies. We complement these algorithmic findings by the following, surprisingly tight lower bound on the asymptotic behavior of our algorithm. We show that unless the Exponential Time Hypothesis (ETH) fails - for any constant 0 <= sigma <= 1, there is p = Theta(k^sigma) such that there is no algorithm deciding in time 2^{o(sqrt{pk})} n^{O(1)} whether an n-vertex graph G can be transformed into a cluster graph with at most p cliques by changing at most k adjacencies. Thus, our upper and lower bounds provide an asymptotically tight analysis of the multivariate parameterized complexity of the problem for the whole range of values of p from constant to a linear function of k.

Integer linear programs (ILPs) are a widely applied framework for dealing with combinatorial problems that arise in practice. It is known, e.g., by the success of CPLEX, that preprocessing and simplification can greatly speed up the process of optimizing an ILP. The present work seeks to further the theoretical understanding of preprocessing for ILPs by initiating a rigorous study within the framework of parameterized complexity and kernelization. A famous result of Lenstra (Mathematics of Operations Research, 1983) shows that feasibility of any ILP with n variables and m constraints can be decided in time O(c^{n^3} m^{c'}). Thus, by a folklore argument, any such ILP admits a kernelization to an equivalent instance of size O(c^{n^3}). It is known, that unless \containment and the polynomial hierarchy collapses, no kernelization with size bound polynomial in n is possible. However, this lower bound only applies for the case when constraints may include an arbitrary number of variables since it follows from lower bounds for \sat and \hittingset, whose bounded arity variants admit polynomial kernelizations. We consider the feasibility problem for ILPs Ax <= b where A is an r-row-sparse matrix parameterized by the number of variables. We show that the kernelizability of this problem depends strongly on the range of the variables. If the range is unbounded then this problem does not admit a polynomial kernelization unless \containment. If, on the other hand, the range of each variable is polynomially bounded in n then we do get a polynomial kernelization. Additionally, this holds also for the more general case when the maximum range d is an additional parameter, i.e., the size obtained is polynomial in n+d.

For an even integer t ≥ 2, the Matching Connectivity matrix Ht is a matrix that has rows and columns both labeled by all perfect matchings of the complete graph Kt on t vertices; an entry Ht[M1,M2] is 1 if M1∪ M2 is a Hamiltonian cycle and 0 otherwise. Motivated by the computational study of the Hamiltonicity problem, we present three results on the structure of Ht: We first show that Ht has rank exactly 2t/2-1 over GF(2) via an appropriate factorization that explicitly provides families of matchings Xt forming bases for Ht. Second, we show how to quickly change representation between such bases. Third, we notice that the sets of matchings Xt induce permutation matrices within Ht. We use the factorization to derive an 1.888n nO(1) time Monte Carlo algorithm that solves the Hamiltonicity problem in directed bipartite graphs. Our algorithm as well counts the number of Hamiltonian cycles modulo two in directed bipartite or undirected graphs in the same time bound. Moreover, we use the fast basis change algorithm from the second result to present a Monte Carlo algorithm that given an undirected graph on n vertices along with a path decomposition of width at most pw decides Hamiltonicity in (2+√2)pw nO(1) time. Finally, we use the third result to show that for every ε >0 this cannot be improved to (2+√2-ε)pwnO(1) time unless the Strong Exponential Time Hypothesis fails, i.e., a faster algorithm for this problem would imply the breakthrough result of an O((2-ε')n) time algorithm for CNF-Sat.

In this paper, we consider multi-objective evolutionary algorithms for the Vertex Cover problem in the context of parameterized complexity. We consider two different measures for the problem. The first measure is a very natural multi-objective one for the use of evolutionary algorithms and takes into account the number of chosen vertices and the number of edges that remain uncovered. The second fitness function is based on a linear programming formulation and proves to give better results. We point out that both approaches lead to a kernelization for the Vertex Cover problem. Based on this, we show that evolutionary algorithms solve the vertex cover problem efficiently if the size of a minimum vertex cover is not too large, i.e., the expected runtime is bounded by O(f(OPT)⋅n c ), where c is a constant and f a function that only depends on OPT. This shows that evolutionary algorithms are randomized fixed-parameter tractable algorithms for the vertex cover problem.

The notion of treewidth plays an important role in theoretical and practical studies of graph problems. It has been recognized that, especially in practical environments, when computing the treewidth of a graph it is invaluable to first apply an array of preprocessing rules that simplify and shrink it. This work seeks to prove rigorous performance guarantees for such preprocessing rules---known rules as well as more recent ones---by studying them in the framework of kernelization from parameterized complexity. It is known that the NP-complete problem of determining whether a given graph $G$ has treewidth at most $k$ admits no polynomial-time preprocessing algorithm that reduces any input instance to size polynomial in $k$, unless NP $\subseteq$ coNP/poly and the polynomial hierarchy collapses to its third level. In this paper we therefore consider structural graph measures larger than treewidth, and determine whether efficient preprocessing can shrink the instance size to a polynomial in such a parameter value. We prove that, given an instance $(G,k)$ of treewidth, we can efficiently reduce its size to $\mathcal{O}(\mathrm{\textsc{fvs}}(G)^4)$ vertices, where $\mathrm{\textsc{fvs}}(G)$ is the size of a minimum feedback vertex set in $G$. We can also prove a size reduction to $\mathcal{O}(\mathrm{\textsc{vc}}(G)^3)$ vertices, where $\mathrm{\textsc{vc}}(G)$ is the size of a minimum vertex cover. Phrased in the language of parameterized complexity, we show that Treewidth has a polynomial kernel when parameterized by the size of a given feedback vertex set, and also by the size of a vertex cover. In contrast we show that Treewidth parameterized by the vertex-deletion distance to a single clique and Weighted Treewidth parameterized by the size of a vertex cover do not admit polynomial kernelizations unless NP $\subseteq$ coNP/poly.

We extend work by Christian et al. [Review of Economic Design 2007] on lobbying in multiple referenda by first providing a more fine-grained analysis of the computational complexity of the NP-complete Lobbying problem. Herein, given a binary matrix, the columns represent issues to vote on and the rows correspond to voters making a binary vote on each issue. An issue is approved if a majority of votes has a 1 in the corresponding column. The goal is to get all issues approved by modifying a minimum number of rows to all-1-rows. In our multivariate complexity analysis, we present a more holistic view on the nature of the computational complexity of Lobbying, providing both (parameterized) tractability and intractability results, depending on various problem parameterizations to be adopted. Moreover, we show non-existence results concerning efficient and effective preprocessing for Lobbying and introduce natural variants such as Restricted Lobbying and Partial Lobbying.

The existence of a polynomial kernel for Odd Cycle Transversal was a notorious open problem in parameterized complexity. Recently, this was settled by the present authors (Kratsch and Wahlstrom, SODA 2012), with a randomized polynomial kernel for the problem, using matroid theory to encode How questions over a set of terminals in size polynomial in the number of terminals (rather than the total graph size, which may be superpolynomially larger). In the current work we further establish the usefulness of matroid theory to kernelization by showing applications of a result on representative sets due to Lovasz (Combinatorial Surveys 1977) and Marx (TCS 2009). We show how representative sets can be used to give a polynomial kernel for the elusive Almost 2-sat problem (where the task is to remove at most k clauses to make a 2-CNF formula satisfiable), solving a major open problem in kernelization. We further apply the representative sets tool to the problem of finding irrelevant vertices in graph cut problems, that is, vertices which can be made undeletable without affecting the status of the problem. This gives the first significant progress towards a polynomial kernel for the Multiway Cut problem; in particular, we get a polynomial kernel for Multiway Cut instances with a bounded number of terminals. Both these kernelization results have significant spin-off effects, producing the first polynomial kernels for a range of related problems. More generally, the irrelevant vertex results have implications for covering min-cuts in graphs. In particular, given a directed graph and a set of terminals, we can find a set of size polynomial in the number of terminals (a cut-covering set) which contains a minimum vertex cut for every choice of sources and sinks from the terminal set. Similarly, given an undirected graph and a set of terminals, we can find a set of vertices, of size polynomial in the number of terminals, which contains a minimum multiway cut for every partition of the terminals into a bounded number of sets. Both results are polynomial time. We expect this to have further applications; in particular, we get direct, reduction rule-based kernelizations for all problems above, in contrast to the indirect compression-based kernel previously given for Odd Cycle Transversal. All our results are randomized, with failure probabilities which can be made exponentially small in the size of the input, due to needing a representation of a matroid to apply the representative sets tool.

Until recently, techniques for obtaining lower bounds for kernelization were one of the most sought after tools in the field of parameterized complexity. Now, after a strong influx of techniques, we are in the fortunate situation of having tools available that are even stronger than what has been required in their applications so far. Based on a result of Fortnow and Santhanam (STOC 2008, JCSS 2011), Bodlaender et al. (ICALP 2008, JCSS 2009) showed that, unless NP ⊆ coNP/poly, the existence of a deterministic polynomial-time composition algorithm, i.e., an algorithm which outputs an instance of bounded parameter value which is yes if and only if one of t input instances is yes, rules out the existence of polynomial kernels for a problem. Dell and van Melkebeek (STOC 2010) continued this line of research and, amongst others, were able to rule out kernels of size O(kd − ∊) for certain problems, assuming NP ⊆ coNP/poly. It is an immediate consequence of their work that even the existence of a co-nondeterministic composition rules out polynomial kernels. However, in contrast to the numerous applications of deterministic composition, the added power of co-nondeterminism has not yet been harnessed to obtain kernelization lower bounds. In this work we present the first example of how co-nondeterminism can help to make a composition algorithm. We study the existence of polynomial kernels for a Ramsey-type problem: Given a graph G and an integer k, the question is whether G contains an independent set or a clique of size at least k. It was asked by Rod Downey whether this problem admits a polynomial kernelization, and such a result would greatly speed up the computation of Ramsey numbers. We provide a co-nondeterministic composition based on embedding t instances into a single host graph H. The crux is that the host graph H needs to observe a bound of ℓ ∊ O(log t) on both its maximum independent set and maximum clique size, while also having a cover of its vertex set by independent sets and cliques all of size ℓ; the co-nondeterministic composition is build around the search for such graphs. Thus we show that, unless NP ⊆ coNP/poly (and the polynomial hierarchy collapses), the problem does not admit a kernelization with polynomial size guarantee.

The Odd Cycle Transversal problem (OCT) asks whether a given graph can be made bipartite by deleting at most k of its vertices. In a breakthrough result Reed, Smith, and Vetta (Operations Research Letters, 2004) gave a O(4k kmn) time algorithm for it, the first algorithm with polynomial runtime of uniform degree for every fixed k. It is known that this implies a polynomial-time compression algorithm that turns OCT instances into equivalent instances of size at most O(4k), a so-called kernelization. Since then the existence of a polynomial kernel for OCT, i.e., a kernelization with size bounded polynomially in k, has turned into one of the main open questions in the study of kernelization. Despite the impressive progress in the area, including the recent development of lower bound techniques (Bodlaender et al., ICALP 2008; Fortnow and Santhanam, STOC 2008) and meta-results on kernelizations for graph problems on planar and other sparse graph classes (Bodlaender et al., FOCS 2009; Fomin et al., SODA 2010), the existence of a polynomial kernel for OCT has remained open, even when the input is restricted to be planar. This work provides the first (randomized) polynomial kernelization for OCT. We introduce a novel kernelization approach based on matroid theory, where we encode all relevant information about a problem instance into a matroid with a representation of size polynomial in k. For OCT, the matroid is built to allow us to simulate the computation of the iterative compression step of the algorithm of Reed, Smith, and Vetta, applied (for only one round) to an approximate odd cycle transversal which it is aiming to shrink to size k. The process is randomized with one-sided error exponentially small in k, where the result can contain false positives but no false negatives, and the size guarantee is cubic in the size of the approximate solution. Combined with an -approximation (Agarwal et al., STOC 2005), we get a reduction of the instance to size O(k4.5), implying a randomized polynomial kernelization. Interestingly, the known lower bound techniques can be seen to exclude randomized kernels that produce no false negatives, as in fact they exclude even co-nondeterministic kernels (Dell and van Melkebeek, STOC 2010). Therefore, our result also implies that deterministic kernels for OCT cannot be excluded by the known machinery.

It has been observed in many places that constant-factor approximable problems often admit polynomial or even linear problem kernels for their decision versions, e.g., Vertex Cover , Feedback Vertex Set , and Triangle Packing . While there exist examples like Bin Packing , which does not admit any kernel unless P = NP, there apparently is a strong relation between these two polynomial-time techniques. We add to this picture by showing that the natural decision versions of all problems in two prominent classes of constant-factor approximable problems, namely MIN F + Π 1 and MAX NP, admit polynomial problem kernels. Problems in MAX SNP, a subclass of MAX NP, are shown to admit kernels with a linear base set, e.g., the set of vertices of a graph. This extends results of Cai and Chen (J. Comput. Syst. Sci. 54(3): 465–474, 1997 ), stating that the standard parameterizations of problems in MAX SNP and MIN F + Π 1 are fixed-parameter tractable, and complements recent research on problems that do not admit polynomial kernelizations (Bodlaender et al. in J. Comput. Syst. Sci. 75(8): 423–434, 2009 ).

We introduce a new technique for proving kernelization lower bounds, called cross-composition. A classical problem L cross-composes into a parameterized problem $Q$ if an instance of Q with polynomially bounded parameter value can express the logical OR of a sequence of instances of L. Building on work by Bodlaender et al. (ICALP 2008) and using a result by Fortnow and Santhanam (STOC 2008) we show that if an NP-hard problem cross-composes into a parameterized problem Q then Q does not admit a polynomial kernel unless the polynomial hierarchy collapses. Our technique generalizes and strengthens the recent techniques of using OR-composition algorithms and of transferring the lower bounds via polynomial parameter transformations. We show its applicability by proving kernelization lower bounds for a number of important graphs problems with structural (non-standard) parameterizations, e.g., Chromatic Number, Clique, and Weighted Feedback Vertex Set do not admit polynomial kernels with respect to the vertex cover number of the input graphs unless the polynomial hierarchy collapses, contrasting the fact that these problems are trivially fixed-parameter tractable for this parameter. We have similar lower bounds for Feedback Vertex Set.

In this paper, we consider multi-objective evolutionary algorithms for the Vertex Cover problem in the context of parameterized complexity. We relate the runtime of our algorithms to the input size and the cost of a minimum solution and point out that the search process of evolutionary algorithms creates partial solutions that are similar to the effect of a kernelization (i.e. a special type of preprocessing from parameterized complexity). Based on this, we show that evolutionary algorithms solve the vertex cover problem efficiently if the size of a minimum vertex cover is not too large, i.e. the expected runtime is bounded by O(f(OPT) nc), where c is a constant and f a function that only depends on OPT. This shows that evolutionary algorithms are randomized fixed-parameter tractable algorithms for the vertex cover problem.

The relation of constant-factor approximability to fixed-parameter tractability and kernelization is a long-standing open question. We prove that two large classes of constant-factor approximable problems, namely~$\textsc{MIN F}^+\Pi_1$ and~$\textsc{MAX NP}$, including the well-known subclass~$\textsc{MAX SNP}$, admit polynomial kernelizations for their natural decision versions. This extends results of Cai and Chen (JCSS 1997), stating that the standard parameterizations of problems in~$\textsc{MAX SNP}$ and~$\textsc{MIN F}^+\Pi_1$ are fixed-parameter tractable, and complements recent research on problems that do not admit polynomial kernelizations (Bodlaender et al.\ ICALP 2008).