arXiv subjects

Markus Sinnl

Publications and source records attributed to Markus Sinnl.

26 records · Page 2Linked to original sources

arXiv2021

Experiments with graph convolutional networks for solving the vertex $p$-center problem

In the last few years, graph convolutional networks (GCN) have become a popular research direction in the machine learning community to tackle NP-hard combinatorial optimization problems (COPs) defined on graphs. While the obtained results are usually still not competitive with problem-specific solution approaches from the operations research community, GCNs often lead to improvements compared to previous machine learning approaches for classical COPs such as the traveling salesperson problem (TSP). In this work we present a preliminary study on using GCNs for solving the vertex p-center problem (PCP), which is another classic COP on graphs. In particular, we investigate whether a successful model based on end-to-end training for the TSP can be adapted to a PCP, which is defined on a similar 2D Euclidean graph input as the usually used version of the TSP. However, the objective of the PCP has a min-max structure which could lead to many symmetric optimal, i.e., ground-truth solutions and other potential difficulties for learning. Our obtained preliminary results show that indeed a direct transfer of network architecture ideas does not seem to work too well. Thus we think that the PCP could be an interesting benchmark problem for new ideas and developments in the area of GCNs.

cs.LG↗

arXiv2021

A branch-and-cut algorithm for submodular interdiction games

Many relevant applications from diverse areas such as marketing, wildlife conservation, or defending critical infrastructure can be modeled as interdiction games. In this work, we introduce interdiction games whose objective is a monotone and submodular set function. Given a ground set of items, the leader interdicts the usage of some of the items of the follower in order to minimize the objective value achievable by the follower, who seeks to maximize a submodular set function over the uninterdicted items subject to knapsack constraints. We propose an exact branch-and-cut algorithm for these kind of interdiction games. The algorithm is based on interdiction cuts which allow to capture the follower's objective function value for a given interdiction decision of the leader and exploit the submodularity of the objective function. We also present extensions and liftings of these cuts and discuss additional preprocessing procedures. We test our solution framework on the weighted maximal covering interdiction game and the bipartite inference interdiction game. For both applications, the improved variants of our interdiction cut perform significantly better than its basic version. For the weighted maximal covering interdiction game for which a mixed-integer bilevel linear programming (MIBLP) formulation is available, we compare the results with those of a state-of-the-art MIBLP solver. While the MIBLP solver yields a minimum of 54% optimality gap within one hour, our best branch-and-cut setting solves all but 4 of 108 instances to optimality with a maximum of 3% gap among unsolved ones.

math.OC↗

arXiv2021

A LP relaxation based matheuristic for multi-objective integer programming

Motivated by their success in the single-objective domain, we propose a very simple linear programming-based matheuristic for tri-objective binary integer programming. To tackle the problem, we obtain lower bound sets by means of the vector linear programming solver Bensolve. Then, simple heuristic approaches, such as rounding and path relinking, are applied to this lower bound set to obtain high-quality approximations of the optimal set of trade-off solutions. The proposed algorithm is compared to a recently suggested algorithm which is, to the best of our knowledge, the only existing matheuristic method for tri-objective integer programming. Computational experiments show that our method produces a better approximation of the true Pareto front using significantly less time than the benchmark method on standard benchmark instances for the three-objective knapsack problem.

math.OC↗

arXiv2019

Exact and heuristic algorithms for the weighted total domination problem

Dominating set problems are among the most important class of combinatorial problems in graph optimization, from a theoretical as well as from a practical point of view. In this paper, we address the recently introduced (minimum) weighted total domination problem. In this problem, we are given an undirected graph with a vertex weight function and an edge weight function. The goal is to find a total dominating set D in this graph with minimal weight. A total dominating set D is a subset of the vertices such that every vertex in the graph, including vertices in D, is adjacent to a vertex in D. The weight is measured as the sum of all vertex weights of vertices in D, plus the edge weights in the subgraph induced by D, plus for each vertex not in D the minimum weight of an edge from it to a vertex in D. In this paper, we present two new Mixed-Integer Programming models for the problem, and design solution frameworks based on them. These solution frameworks also include valid inequalities, starting heuristics and primal heuristics. In addition, we also develop a genetic algorithm, which is based on a greedy randomized adaptive search procedure version of our starting heuristic. We carry out a computational study to assess the performance of our approaches when compared to the previous work for the same problem. The study reveals that our exact solution algorithms are up to 500 times faster compared to previous exact approaches and instances with up to 125 vertices can be solved to optimality within a timelimit of 1800 seconds. Moreover, the presented genetic algorithm also works well and often finds the optimal or a near-optimal solution within a short runtime. Additionally, we also analyze the influence of instance-characteristics on the performance of our algorithms.

math.OC↗

arXiv2019

A note on computational approaches for the antibandwidth problem

In this note, we consider the antibandwidth problem, also known as dual bandwidth problem, separation problem and maximum differential coloring problem. Given a labeled graph (i.e., a numbering of the vertices of a graph), the antibandwidth of a node is defined as the minimum absolute difference of its labeling to the labeling of all its adjacent vertices. The goal in the antibandwidth problem is to find a labeling maximizing the antibandwidth. The problem is NP-hard in general graphs and has applications in diverse areas like scheduling, radio frequency assignment, obnoxious facility location and map-coloring. There has been much work on deriving theoretical bounds for the problem and also in the design of metaheuristics in recent years. However, the optimality gaps between the best known solution values and reported upper bounds for the HarwellBoeing Matrix-instances, which are the commonly used benchmark instances for this problem, are often very large (e.g., up to 577%). The upper bounds reported in literature are based on the theoretical bounds involving simple graph characteristics, i.e., size, order and degree, and a mixed-integer programming (MIP) model. We present new MIP models for the problem, together with valid inequalities, and design a branch-and-cut algorithm and an iterative solution algorithm based on them. These algorithms also include two starting heuristics and a primal heuristic. We also present a constraint programming approach, and calculate upper bounds based on the stability number and chromatic number. Our computational study shows that the developed approaches allow to find the proven optimal solution for eight instances from literature, where the optimal solution was unknown and also provide reduced gaps for eleven additional instances, including improved solution values for seven instances, the largest optimality gap is now 46%.

cs.DS↗

arXiv2019

An exact solution framework for the multiple gradual cover location problem

Facility and covering location models are key elements in many decision aid tools in logistics, supply chain design, telecommunications, public infrastructure planning, and many other industrial and public sectors. In many applications, it is likely that customers are not dichotomously covered by facilities, but gradually covered according to, e.g., the distance to the open facilities. Moreover, customers are not served by a single facility, but by a collection of them, which jointly serve them. In this paper we study the recently introduced multiple gradual cover location problem (MGCLP). The MGCLP addresses both of the issues described above. We provide four different mixed-integer programming formulations for the MGCLP, all of them exploiting the submodularity of the objective function and developed a branch-and-cut framework based one these formulations. The framework is further enhanced by starting and primal heuristics and initialization procedures. The computational results show that our approach allows to effectively address different sets of instances. We provide optimal solution values for 13 instances from literature, where the optimal solution was not known, and additionally provide improved solution values for seven instances. Many of these instances can be solved within a minute. We also analyze the dependence of the solution-structure on instance-characteristics.

math.OC↗

arXiv2019

Algorithmic expedients for the S-labeling problem

Graph labeling problems have been widely studied in the last decades and have a vast area of application. In this work, we study the recently introduced S-labeling problem, in which the nodes get labeled using labels from 1 to |V | and for each edge the contribution to the objective function, called S-labeling number of the graph, is the minimum label of its end-nodes. The goal is to find a labeling with minimum value. The problem is NP-hard for planar subcubic graphs, although for many other graph classes the complexity status is still unknown. In this paper, we present different algorithmic approaches for tackling this problem: We develop an exact solution framework based on Mixed-Integer Programming (MIP) which is enhanced with valid inequalities, starting and primal heuristics and specialized branching rules. We show that our MIP formulation has no integrality gap for paths, cycles and perfect n-ary trees, and, to the best of our knowledge, we give the first polynomial-time algorithm for the problem on n-ary trees as well as a closed formula for the S-labeling number for such trees. Moreover, we also present a Lagrangian heuristic and a constraint programming approach. A computational study is carried out in order to (i) investigate if there may be other special graph classes, where our MIP formulation has no integrality gap, and (ii) assess the effectiveness of the proposed solution approaches for solving the problem on a dataset consisting of general graphs.

math.OC↗

arXiv2019

The Generalized Reserve Set Covering Problem with Connectivity and Buffer Requirements

The design of nature reserves is becoming, more and more, a crucial task for ensuring the conservation of endangered wildlife. In order to guarantee the preservation of species and a general ecological functioning, the designed reserves must typically verify a series of spatial requirements. Among the required characteristics, practitioners and researchers have pointed out two crucial aspects: (i) connectivity, so as to avoid spatial fragmentation, and (ii) the design of buffer zones surrounding (or protecting) so-called core areas. In this paper, we introduce the Generalized Reserve Set Covering Problem with Connectivity and Buffer Requirements. This problem extends the classical Reserve Set Covering Problem and allows to address these two requirements simultaneously. A solution framework based on Integer Linear Programming and branch-and-cut is developed. The framework is enhanced by valid inequalities, a construction and a primal heuristic and local branching. An extensive computational study on grid-graph instances and real-life instances based on data from three states of the U.S. and one region of Australia is carried out to assess the suitability of the proposed model to deal with the challenges faced by decision-makers in natural reserve design. The results show, on the one hand, the flexibility of the proposed models to provide solutions according to the decision-makers' requirements, and on the other hand, the effectiveness of the devised algorithm for providing' good solutions in reasonable computing times.

math.OC↗