Clean Citation Style
{ "authors" : [{ "lastname":"Bläsius" , "initial":"T" , "url":"https://hpi.de/friedrich/publications/people/thomasblaesius.html" , "mail":"thomas.blasius(at)hpi.de" }, { "lastname":"Casel" , "initial":"K" , "url":"https://hpi.de/friedrich/publications/people/katrincasel.html" , "mail":"katrin.casel(at)hpi.de" }, { "lastname":"Chauhan" , "initial":"A" , "url":"https://hpi.de/friedrich/publications/people/ankitchauhan.html" , "mail":"ankit.chauhan(at)hpi.de" }, { "lastname":"Doskoč" , "initial":"V" , "url":"https://hpi.de/friedrich/publications/people/vanjadoskoc.html" , "mail":"vanja.doskoc(at)hpi.de" }, { "lastname":"Fischbeck" , "initial":"P" , "url":"https://hpi.de/friedrich/publications/people/philippfischbeck.html" , "mail":"philipp.fischbeck(at)hpi.de" }, { "lastname":"Friedrich" , "initial":"T" , "url":"https://hpi.de/friedrich/publications/people/tobiasfriedrich.html" , "mail":"friedrich(at)hpi.de" }, { "lastname":"Göbel" , "initial":"A" , "url":"https://hpi.de/friedrich/publications/people/andreasgoebel.html" , "mail":"andreas.goebel(at)hpi.de" }, { "lastname":"Katzmann" , "initial":"M" , "url":"https://hpi.de/friedrich/publications/people/maximiliankatzmann.html" , "mail":"maximilian.katzmann(at)hpi.de" }, { "lastname":"Kötzing" , "initial":"T" , "url":"https://hpi.de/friedrich/publications/people/timokoetzing.html" , "mail":"timo.koetzing(at)hpi.de" }, { "lastname":"Krejca" , "initial":"M" , "url":"https://hpi.de/friedrich/publications/people/martinkrejca.html" , "mail":"martin.krejca(at)hpi.de" }, { "lastname":"Krohmer" , "initial":"A" , "url":"https://hpi.de/friedrich/publications/people/antonkrohmer.html" , "mail":"none" }, { "lastname":"Lagodzinski" , "initial":"G" , "url":"https://hpi.de/friedrich/publications/people/gregorlagodzinski.html" , "mail":"gregor.lagodzinski(at)hpi.de" }, { "lastname":"Lenzner" , "initial":"P" , "url":"https://hpi.de/friedrich/publications/people/pascallenzner.html" , "mail":"pascal.lenzner(at)hpi.de" }, { "lastname":"Melnichenko" , "initial":"A" , "url":"https://hpi.de/friedrich/publications/people/annamelnichenko.html" , "mail":"anna.melnichenko(at)hpi.de" }, { "lastname":"Molitor" , "initial":"L" , "url":"https://hpi.de/friedrich/publications/people/louisemolitor.html" , "mail":"louise.molitor(at)hpi.de" }, { "lastname":"Neubert" , "initial":"S" , "url":"https://hpi.de/friedrich/publications/people/stefanneubert.html" , "mail":"stefan.neubert(at)hpi.de" }, { "lastname":"Quinzan" , "initial":"F" , "url":"https://hpi.de/friedrich/publications/people/francescoquinzan.html" , "mail":"francesco.quinzan(at)hpi.de" }, { "lastname":"Rothenberger" , "initial":"R" , "url":"https://hpi.de/friedrich/publications/people/ralfrothenberger.html" , "mail":"ralf.rothenberger(at)hpi.de" }, { "lastname":"Schirneck" , "initial":"M" , "url":"https://hpi.de/friedrich/publications/people/martinschirneck.html" , "mail":"martin.schirneck(at)hpi.de" }, { "lastname":"Seidel" , "initial":"K" , "url":"https://hpi.de/friedrich/publications/people/karenseidel.html" , "mail":"karen.seidel(at)hpi.de" }, { "lastname":"Sutton" , "initial":"A" , "url":"https://hpi.de/friedrich/publications/people/andrewmsutton.html" , "mail":"none" }]}

Shi, Feng; Schirneck, Martin; Friedrich, Tobias; Kötzing, Timo; Neumann, Frank Erratum to: Reoptimization Time Analysis of Evolutionary Algorithms on Linear Functions Under Dynamic Uniform Constraints. Algorithmica 2018
In the article "Reoptimization Time Analysis of Evolutionary Algorithms on Linear Functions Under Dynamic Uniform Constraints", we claimed a worstcase runtime of \(O(n D \log D)\) and \(O(n D)\) for the MultiObjective Evolutionary Algorithm and the MultiObjective \((μ+(λ,λ))\) Genetic Algorithm, respectively, on linear profit functions under dynamic uniform constraint. The technique used to prove these results c contained an error. Instead, we correct this mistake and show a weaker bound of \(O(n D^2)\) for both algorithms.

Doerr, Benjamin; Fischbeck, Philipp; Frahnow, Clemens; Friedrich, Tobias; Kötzing, Timo; Schirneck, Martin Island Models Meet Rumor Spreading. Algorithmica 2018
Island models in evolutionary computation solve problems by a careful interplay of independently running evolutionary algorithms on the island and an exchange of good solutions between the islands. In this work, we conduct rigorous run time analyses for such island models trying to simultaneously obtain good run times and low communication effort. We improve the existing upper bounds for both measures (i) by improving the run time bounds via a careful analysis, (ii) by balancing individual computation and communication in a more appropriate manner, and (iii) by replacing the usual communicatewithall approach with randomized rumor spreading. In the latter, each island contacts a randomly chosen neighbor. This epidemic communication paradigm is known to lead to very fast and robust information dissemination in many applications. Our results concern island models running simple (1+1) evolutionary algorithms to optimize the classic test functions OneMax and LeadingOnes. We investigate binary trees, ddimensional tori, and complete graphs as communication topologies.

Shi, Feng; Schirneck, Martin; Friedrich, Tobias; Kötzing, Timo; Neumann, Frank Reoptimization Time Analysis of Evolutionary Algorithms on Linear Functions Under Dynamic Uniform Constraints. Algorithmica 2018
Rigorous runtime analysis is a major approach towards understanding evolutionary computing techniques, and in this area linear pseudoBoolean objective functions play a central role. Having an additional linear constraint is then equivalent to the NPhard Knapsack problem, certain classes thereof have been studied in recent works. In this article, we present a dynamic model of optimizing linear functions under uniform constraints. Starting from an optimal solution with respect to a given constraint bound, we investigate the runtimes that different evolutionary algorithms need to recompute an optimal solution when the constraint bound changes by a certain amount. The classical \((1+1)\) EA and several populationbased algorithms are designed for that purpose, and are shown to recompute efficiently. Furthermore, a variant of the \((1+(λ,λ))\) GA for the dynamic optimization problem is studied, whose performance is better when the change of the constraint bound is small.

Bläsius, Thomas; Karrer, Annette; Rutter, Ignaz Simultaneous Embedding: Edge Orderings, Relative Positions, Cutvertices. Algorithmica 2018: 12141277
A simultaneous embedding (with fixed edges) of two graphs \(G^1\) and \(G^2\) with common graph \(G = G^1 ∩ G^2\) is a pair of planar drawings of \(G^1\) and \(G^2\) that coincide on \(G\). It is an open question whether there is a polynomialtime algorithm that decides whether two graphs admit a simultaneous embedding (problem SEFE). In this paper, we present two results. First, a set of three lineartime preprocessing algorithms that remove certain substructures from a given SEFE instance, producing a set of equivalent Sefe instances without such substructures. The structures we can remove are (1) cutvertices of the union graph \(G^∪ = G^1 ∪ G^2\) , (2) most separating pairs of \(G^∪\), and (3) connected components of G that are biconnected but not a cycle. Second, we give an \(O(n³)\)time algorithm solving Sefe for instances with the following restriction. Let u be a pole of a Pnode \(µ\) in the SPQRtree of a block of \(G^1\) or \(G^2\). Then at most three virtual edges of \(µ\) may contain common edges incident to u. All algorithms extend to the sunflower case, i.e., to the case of more than two graphs pairwise intersecting in the same common graph.

Kötzing, Timo; Sudholt, Dirk Preface to the Special Issue on Theory of Genetic and Evolutionary Computation. Algorithmica 2018: 15751578
Evolutionary algorithms (EAs) are randomized search heuristics that can be employed to solve complex optimization problems, including multimodal or highly constrained problems. EAs work by mimicking principles from natural evolution: maintaining a collection of possible solutions (a population) and iteratively creating variants of the individuals (the offspring) and then choosing a new set of individuals for the next iteration (selection). EAs are popular because they represent generalpurpose optimizers that can be easily applied to various problems, even in cases where little or no indepth knowledge about the problem is available. In order to guide practitioners devising new and effective algorithms, theoretical computer scientists employ methods from the field of randomized algorithms to analyze the working principles of EAs with mathematical rigor. Key questions concern the impact of parameter choices (such as, for example, the offspring size or the choice of variation operators) as well as foundational work on developing powerful analysis methods. The theory track of the annual ACM Genetic and Evolutionary Computation Conference (GECCO) is the first tier event for advances in this direction. In this special issue six selected papers from the 2016 edition of the GECCO theory track are collected, each one of them carefully revised and extended to meet the high quality standards of Algorithmica.

Doerr, Benjamin; Doerr, Carola; Kötzing, Timo Static and SelfAdjusting Mutation Strengths for Multivalued Decision Variables. Algorithmica 2018: 17321768
The most common representation in evolutionary computation are bit strings. With very little theoretical work existing on how to use evolutionary algorithms for decision variables taking more than two values, we study the run time of simple evolutionary algorithms on some OneMaxlike functions defined over \(\Omega={0,1,\dots,r−1}^n\). We observe a crucial difference in how we extend the onebitflip and standardbit mutation operators to the multivalued domain. While it is natural to modify a random position of the string or select each position of the solution vector for modification independently with probability \(1/n\), there are various ways to then change such a position. If we change each selected position to a random value different from the original one, we obtain an expected run time of \(\Theta(n r \log n)\). If we change each selected position by \(+1\) or \(−1\) (random choice), the optimization time reduces to \(\Theta(n r + n \log n)\). If we use a random mutation strength \(i \in {0,1,\dots,r−1}\) with probability inversely proportional to \(i\) and change the selected position by \(+i\) or \(−i\) (random choice), then the optimization time becomes \(\Theta(n \log(r)(\log n + \log r))\), which is asymptotically faster than the previous if \(r = \omega(\log(n)\log\log(n))\). Interestingly, a better expected performance can be achieved with a selfadjusting mutation strength that is based on the success of previous iterations. For the mutation operator that modifies a randomly chosen position, we show that the selfadjusting mutation strength yields an expected optimization time of \(\Theta(n(\log n + \log r))\), which is best possible among all dynamic mutation strengths. In our proofs, we use a new multiplicative drift theorem for computing lower bounds, which is not restricted to processes that move only towards the target.

Bläsius, Thomas; Friedrich, Tobias; Krohmer, Anton Cliques in Hyperbolic Random Graphs. Algorithmica 2018: 23242344
Most complex real world networks display scalefree features. This characteristic motivated the study of numerous random graph models with a powerlaw degree distribution. There is, however, no established and simple model which also has a high clustering of vertices as typically observed in real data. Hyperbolic random graphs bridge this gap. This natural model has recently been introduced by hyprg and has shown theoretically and empirically to fulfill all typical properties of real world networks, including powerlaw degree distribution and high clustering. We study cliques in hyperbolic random graphs \(G\) and present new results on the expected number of \(k\)cliques \(E(K_k)\) and the size of the largest clique \(\omega(G)\). We observe that there is a phase transition at powerlaw exponent \(\beta = 3\). More precisely, for \(beta in (2,3)\) we prove \(E(K_k)=n^k (3\beta)/2} \Theta(k)^{k}\) and \(\omega(G)=\Theta(n^(3\beta)/2})\), while for \(\beta\geq3\) we prove \(E(K_k)=n \Theta(k)^{k}\) and \(\omega(G)=\Theta(\log(n)/ \log \log n)\). Furthermore, we show that for \(\beta \geq 3\), cliques in hyperbolic random graphs can be computed in time \(O(n)\). If the underlying geometry is known, cliques can be found with worstcase runtime \(O(m \cdot n^{2.5})\) for all values of \(\beta\).

AbuKhzam, Faisal N.; Bazgan, Cristina; Casel, Katrin; Fernau, Henning Clustering with LowerBounded Sizes  A General GraphTheoretic Framework. Algorithmica 2018: 25172550
Classical clustering problems search for a partition of objects into a fixed number of clusters. In many scenarios, however, the number of clusters is not known or necessarily fixed. Further, clusters are sometimes only considered to be of significance if they have a certain size. We discuss clustering into sets of minimum cardinality \(k\) without a fixed number of sets and present a general model for these types of problems. This general framework allows the comparison of different measures to assess the quality of a clustering. We specifically consider nine qualitymeasures and classify the complexity of the resulting problems with respect to \(k\). Further, we derive some polynomialtime solvable cases for \(k=2\) with connections to matchingtype problems which, among other graph problems, then are used to compute approximations for larger values of \(k\).

Bringmann, Karl; Friedrich, Tobias; Krohmer, Anton Deanonymization of Heterogeneous Random Graphs in Quasilinear Time. Algorithmica 2018: 3397–3427
There are hundreds of online social networks with altogether billions of users. Many such networks publicly release structural information, with all personal information removed. Empirical studies have shown, however, that this provides a false sense of privacy  it is possible to identify almost all users that appear in two such anonymized network as long as a few initial mappings are known. We analyze this problem theoretically by reconciling two versions of an artificial powerlaw network arising from independent subsampling of vertices and edges. We present a new algorithm that identifies most vertices and makes no wrong identifications with high probability. The number of vertices matched is shown to be asymptotically optimal. For an nvertex graph, our algorithm uses \(n^\varepsilon\) seed nodes (for an arbitrarily small \(\varepsilon\)) and runs in quasilinear time. This improves previous theoretical results which need \(\Theta(n)\) seed nodes and have runtimes of order \(n^{1+\Omega(1)}\). Additionally, the applicability of our algorithm is studied experimentally on different networks.

Aschenbach, Martin; Kötzing, Timo; Seidel, Karen Learning from Informants: Relations between Learning Success Criteria. ArXiv 2018: 37
Learning from positive and negative information, socalled informants, being one of the models for human and machine learning introduced by Gold [1967], is investigated. Particularly, naturally arising questions about this learning setting, originating in results on learning from solely positive information, are answered. By a carefully arranged argument learners can be assumed to only change their hypothesis in case it is inconsistent with the data (such a learning behavior is called conservative). The deduced main theorem states the relations between the most important delayable learning success criteria, being the ones not ruined by a delayed in time hypothesis output. Additionally, our investigations concerning the nondelayable requirement of consistent learning underpin the claim for delayability being the right structural property to gain a deeper understanding concerning the nature of learning success criteria. Moreover, we obtain an anomalous hierarchy when allowing for an increasing finite number of anomalies of the hypothesized language by the learner compared with the language to be learned. In contrast to the vacillatory hierarchy for learning from solely positive information, we observe a duality depending on whether infinitely many vacillations between different (almost) correct hypotheses are still considered a successful learning behavior.

Bläsius, Thomas; Stumpf, Peter; Ueckerdt, Torsten Local and Union Boxicity. Discrete Mathematics 2018: 1307  1315
The boxicity \(box(H)\) of a graph \(H\) is the smallest integer \(d\) such that \(H\) is the intersection of \(d\) interval graphs, or equivalently, that \(H\) is the intersection graph of axisaligned boxes in \(R^d\). These intersection representations can be interpreted as covering representations of the complement \(H^c\) of \(H\) with cointerval graphs, that is, complements of interval graphs. We follow the recent framework of global, local and folded covering numbers (Knauer and Ueckerdt, 2016) to define two new parameters: the local boxicity \(box_l(H)\) and the union boxicity \(box_u(H)\) of \(H\). The union boxicity of \(H\) is the smallest \(d\) such that \(H^c\) can be covered with \(d\) vertexdisjoint unions of cointerval graphs, while the local boxicity of \(H\) is the smallest \(d\) such that \(H^c\) can be covered with cointerval graphs, at most \(d\) at every vertex. We show that for every graph \(H\) we have \(box_l(H) leq box_u(H) leq box(H) \) and that each of these inequalities can be arbitrarily far apart. Moreover, we show that local and union boxicity are also characterized by intersection representations of appropriate axisaligned boxes in \(R^d\) . We demonstrate with a few striking examples, that in a sense, the local boxicity is a better indication for the complexity of a graph, than the classical boxicity.

Dang, DucCuong; Friedrich, Tobias; Kötzing, Timo; Krejca, Martin S.; Lehre, Per Kristian; Oliveto, Pietro S.; Sudholt, Dirk; Sutton, Andrew M. Escaping Local Optima Using Crossover with Emergent Diversity. IEEE Transactions on Evolutionary Computation 2018: 484497
Population diversity is essential for the effective use of any crossover operator. We compare seven commonly used diversity mechanisms and prove rigorous run time bounds for the \((\mu+1)\) GA using uniform crossover on the fitness function \(Jump_k\). All previous results in this context only hold for unrealistically low crossover probability \(p_c=O(k/n)\), while we give analyses for the setting of constant \(p_c < 1\) in all but one case. Our bounds show a dependence on the problem size \(n\), the jump length \(k\), the population size \(\mu\), and the crossover probability \(p_c\). For the typical case of constant \(k > 2\) and constant \(p_c\), we can compare the resulting expected optimisation times for different diversity mechanisms assuming an optimal choice of \(\mu\): \(O(n^{k1})\) for duplicate elimination/minimisation, \(O(n^2 \log n)\) for maximising the convex hull, \(O(n \log n)\) for det. crowding (assuming \(p_c = k/n\)), \(O(n \log n)\) for maximising the Hamming distance, \(O(n \log n)\) for fitness sharing, \(O(n \log n)\) for the singlereceiver island model. This proves a sizeable advantage of all variants of the \((\mu+1)\) GA compared to the (1+1) EA, which requires \(\Theta(n^k)\). In a short empirical study we confirm that the asymptotic differences can also be observed experimentally.

Bläsius, Thomas; Friedrich, Tobias; Krohmer, Anton; Laue, Sören Efficient Embedding of ScaleFree Graphs in the Hyperbolic Plane. IEEE/ACM Transactions on Networking 2018: 920933
Hyperbolic geometry appears to be intrinsic in many large real networks. We construct and implement a new maximum likelihood estimation algorithm that embeds scalefree graphs in the hyperbolic space. All previous approaches of similar embedding algorithms require at least a quadratic runtime. Our algorithm achieves quasilinear runtime, which makes it the first algorithm that can embed networks with hundreds of thousands of nodes in less than one hour. We demonstrate the performance of our algorithm on artificial and real networks. In all typical metrics, like Loglikelihood and greedy routing, our algorithm discovers embeddings that are very close to the ground truth.

Baum, Moritz; Bläsius, Thomas; Gemsa, Andreas; Rutter, Ignaz; Wegner, Franziska Scalable Exact Visualization of Isocontours in Road Networks via MinimumLink Paths. Journal of Computational Geometry 2018: 2470
Isocontours in road networks represent the area that is reachable from a source within a given resource limit. We study the problem of computing accurate isocontours in realistic, largescale networks. We propose isocontours represented by polygons with minimum number of segments that separate reachable and unreachable components of the network. Since the resulting problem is not known to be solvable in polynomial time, we introduce several heuristics that run in (almost) linear time and are simple enough to be implemented in practice. A key ingredient is a new practical lineartime algorithm for minimumlink paths in simple polygons. Experiments in a challenging realistic setting show excellent performance of our algorithms in practice, computing nearoptimal solutions in a few milliseconds on average, even for long ranges.

Friedrich, Tobias; Katzmann, Maximilian; Krohmer, Anton Unbounded Discrepancy of Deterministic Random Walks on Grids. SIAM Journal on Discrete Mathematics 2018
Random walks are frequently used in randomized algorithms. We study a derandomized variant of a random walk on graphs, called rotorrouter model. In this model, instead of distributing tokens randomly, each vertex serves its neighbors in a fixed deterministic order. For most setups, both processes behave remarkably similar: Starting with the same initial configuration, the number of tokens in the rotorrouter model deviates only slightly from the expected number of tokens on the corresponding vertex in the random walk model. The maximal difference over all vertices and all times is called single vertex discrepancy. Cooper and Spencer (2006) showed that on \(\mathbb{Z}^{d}\) the single vertex discrepancy is only a constant \(c_d\). Other authors also determined the precise value of \(c_d\) for \(d=1,2\). All these results, however, assume that initially all tokens are only placed on one partition of the bipartite graph \(\mathbb{Z}^{d}\). We show that this assumption is crucial by proving that otherwise the single vertex discrepancy can become arbitrarily large. For all dimensions \(d\geq1\) and arbitrary discrepancies~\(\ell \geq 0\), we construct configurations that reach a discrepancy of at least \(\ell\).

Friedrich, Tobias; Krohmer, Anton On the diameter of hyperbolic random graphs. SIAM Journal on Discrete Mathematics 2018: 13141334
Large realworld networks are typically scalefree. Recent research has shown that such graphs are described best in a geometric space. More precisely, the internet can be mapped to a hyperbolic space such that geometric greedy routing is close to optimal (Bogu\~n}{á}, Papadopoulos, and Krioukov. Nature Communications, 1:62, 2010). This observation has pushed the interest in hyperbolic networks as a natural model for scalefree networks. Hyperbolic random graphs follow a power law degree distribution with controllable exponent \(\beta)\) and show high clustering (Gugelmann, Panagiotou, and Peter. ICALP, pp. 573–585, 2012). For understanding the structure of the resulting graphs and for analyzing the behavior of network algorithms, the next question is bounding the size of the diameter. The only known explicit bound is \(O((\log n)^32/((3  \beta)(5  \beta))+1})\)(Kiwi and Mitsche. ANALCO, pp. 26–39, 2015). We present two much simpler proofs for an improved upper bound of \(O((\log n)^2/(3  \beta)})\) and a lower bound of \(\Omega(\log n)\). If \(\beta > 3\), we show that the latter bound is tight by proving an upper bound of \(O(\log n)\) for the diameter.

Friedrich, Tobias; Kötzing, Timo; Lagodzinski, J. A. Gregor; Neumann, Frank; Schirneck, Martin Analysis of the (1+1) EA on Subclasses of Linear Functions under Uniform and Linear Constraints. Theoretical Computer Science 2018
Linear functions have gained great attention in the run time analysis of evolutionary computation methods. The corresponding investigations have provided many effective tools for analyzing more complex problems. So far, the runtime analysis of evolutionary algorithms has mainly focused on unconstrained problems, but problems occurring in applications frequently involve constraints. Therefore, there is a strong need to extend the methods for analyzing unconstrained problems to a setting involving constraints. In this paper, we consider the behavior of the classical (1+1) evolutionary algorithm on linear functions under linear constraint. We show tight bounds in the case where the constraint is given by the OneMax function and the objective function is given by either the OneMax or the BinVal function. For the general case we present upper and lower bounds.

Krejca, Martin; Witt, Carsten Lower Bounds on the Run Time of the Univariate Marginal Distribution Algorithm on OneMax. Theoretical Computer Science 2018
The Univariate Marginal Distribution Algorithm (UMDA)  a popular estimationofdistribution algorithm  is studied from a run time perspective. On the classical OneMax benchmark function on bit strings of length \(n\), a lower bound of \(\Omega(\lambda + \mu \sqrt{n} + n\log n)\), where \(\mu\) and \(\lambda\) are algorithmspecific parameters, on its expected run time is proved. This is the first direct lower bound on the run time of UMDA. It is stronger than the bounds that follow from general blackbox complexity theory and is matched by the run time of many evolutionary algorithms. The results are obtained through advanced analyses of the stochastic change of the frequencies of bit values maintained by the algorithm, including carefully designed potential functions. These techniques may prove useful in advancing the field of run time analysis for estimationofdistribution algorithms in general.

Bazgan, Cristina; Brankovic, Ljiljana; Casel, Katrin; Fernau, Henning; Jansen, Klaus; Klein, KimManuel; Lampis, Michael; Liedloff, Mathieu; Monnot, Jérôme; Paschos, Vangelis Th. The many facets of upper domination. Theoretical Computer Science 2018: 225
This paper studies Upper Domination, i.e., the problem of computing the maximum cardinality of a minimal dominating set in a graph with respect to classical and parameterised complexity as well as approximability.