A hybrid breakout local search and reinforcement learning approach to the vertex separator problem

The Vertex Separator Problem (VSP) is an NP-hard problem which emerges from a variety of important domains and applications. In this paper, we present an improved Breakout Local Search for VSP (named BLS-RLE). The distinguishing feature of this approach is a new parameter control mechanism that draws upon ideas from reinforcement learning theory to reach an interdependent decision on the number and on the type of perturbation moves. The mechanism complies with the principle of first carrying out intensification and then employing minimal diversification only if needed, it uses a dedicated sampling strategy for a rapid convergence towards a limited set of parameter values that appear to be the most convenient for the given state of search. Extensive experimental evaluations and statistical comparisons on a wide range of benchmark instances show significant improvement in performance of the proposed algorithm over the existing BLS algorithm for VSP. Indeed, out of the 422 tested instances, BLS-RLE was able to attain the best-known solution in 93.8% of the cases, which is around 20% higher compared to the existing BLS. In addition, we provide detailed analyses to evaluate the importance of the key elements of the proposed method and to justify the degree of diversification introduced during perturbation.

[1]  Egon Balas,et al.  The vertex separator problem: a polyhedral investigation , 2005, Math. Program..

[2]  Dimitris K. Tasoulis,et al.  Tracking Particle Swarm Optimizers: An adaptive approach through multinomial distribution tracking with exponential forgetting , 2012, 2012 IEEE Congress on Evolutionary Computation.

[3]  Mark Hoogendoorn,et al.  Why parameter control mechanisms should be benchmarked against random variation , 2013, 2013 IEEE Congress on Evolutionary Computation.

[4]  Nicholas Hopper,et al.  Routing around decoys , 2012, CCS.

[5]  Franz Rendl,et al.  A Spectral Bundle Method for Semidefinite Programming , 1999, SIAM J. Optim..

[6]  James R. Lee,et al.  Improved approximation algorithms for minimum-weight vertex separators , 2005, STOC '05.

[7]  Paul Erdös,et al.  On random graphs, I , 1959 .

[8]  D. Wolfe,et al.  Nonparametric Statistical Methods. , 1974 .

[9]  Michael G. Epitropakis,et al.  Evolutionary adaptation of the differential evolution control parameters , 2009, 2009 IEEE Congress on Evolutionary Computation.

[10]  Somayé Ghandi,et al.  A breakout local search (BLS) method for solving the assembly sequence planning problem , 2015, Eng. Appl. Artif. Intell..

[11]  Uriel Feige,et al.  Finding small balanced separators , 2006, STOC '06.

[12]  Jin-Kao Hao,et al.  Breakout Local Search for maximum clique problems , 2013, Comput. Oper. Res..

[13]  Bin Fu,et al.  Sublinear time width-bounded separators and their application to the protein side-chain packing problem , 2006, J. Comb. Optim..

[14]  Egon Balas,et al.  The vertex separator problem: algorithms and computations , 2005, Math. Program..

[15]  Silvestre Fialho,et al.  Adaptive operator selection for optimization , 2010 .

[16]  Massimo Marchiori,et al.  Error and attacktolerance of complex network s , 2004 .

[17]  Steven David Prestwich,et al.  Tuning Local Search by Average-Reward Reinforcement Learning , 2008, LION.

[18]  Richard S. Sutton,et al.  Introduction to Reinforcement Learning , 1998 .

[19]  Mauro Brunato,et al.  Learning While Optimizing an Unknown Fitness Surface , 2008, LION.

[20]  Marie-Jean Meurs,et al.  An exact algorithm for solving the vertex separator problem , 2011, J. Glob. Optim..

[21]  R. Tarjan,et al.  A Separator Theorem for Planar Graphs , 1977 .

[22]  Helena Ramalhinho Dias Lourenço,et al.  Iterated Local Search , 2001, Handbook of Metaheuristics.

[23]  Boulevard Lavoisier,et al.  Breakout local search for the quadratic assignment problem , 2013 .

[24]  Donghyun Kim,et al.  On Combinatoric Approach to Circumvent Internet Censorship Using Decoy Routers , 2013, MILCOM 2013 - 2013 IEEE Military Communications Conference.

[25]  Frank Thomson Leighton,et al.  A Framework for Solving VLSI Graph Layout Problems , 1983, J. Comput. Syst. Sci..

[26]  Ümit V. Çatalyürek,et al.  Partitioning Hypergraphs in Scientific Computing Applications through Vertex Separators on Graphs , 2012, SIAM J. Sci. Comput..

[27]  Martijn C. Schut,et al.  Reinforcement Learning for Online Control of Evolutionary Algorithms , 2006, ESOA.

[28]  Cid C. de Souza,et al.  Exact algorithms for the vertex separator problem in graphs , 2011, Networks.

[29]  William W. Hager,et al.  Continuous quadratic programming formulations of optimization problems on graphs , 2015, Eur. J. Oper. Res..

[30]  Mark Hoogendoorn,et al.  Parameter Control in Evolutionary Algorithms: Trends and Challenges , 2015, IEEE Transactions on Evolutionary Computation.

[31]  Dirk Thierens,et al.  Adaptive Strategies for Operator Allocation , 2007, Parameter Setting in Evolutionary Algorithms.

[32]  Peter Auer,et al.  Finite-time Analysis of the Multiarmed Bandit Problem , 2002, Machine Learning.

[33]  Jin-Kao Hao,et al.  Breakout Local Search for the Vertex Separator Problem , 2013, IJCAI.

[34]  Jesús Sánchez-Oro,et al.  General Variable Neighborhood Search for computing graph separators , 2017, Optim. Lett..

[35]  Charles E. Leiserson,et al.  Area-efficient graph layouts , 1980, 21st Annual Symposium on Foundations of Computer Science (sfcs 1980).

[36]  A. George.  An automatic one-way dissection algorithm for irregular finite element problems , 1978 .

[37]  Albert-László Barabási,et al.  Statistical mechanics of complex networks , 2001, ArXiv.

[38]  Joseph W. H. Liu,et al.  A graph partitioning algorithm by node separators , 1989, TOMS.

[39]  H. Robbins.  Some aspects of the sequential design of experiments , 1952 .

[40]  Admela Jukan,et al.  SDN Partitioning: A Centralized Control Plane for Distributed Routing Protocols , 2016, IEEE Transactions on Network and Service Management.

[41]  Dimitris K. Tasoulis,et al.  Tracking Differential Evolution Algorithms: An Adaptive Approach through Multinomial Distribution Tracking with Exponential Forgetting , 2012, SETN.

[42]  Curt Jones,et al.  Finding Good Approximate Vertex and Edge Partitions is NP-Hard , 1992, Inf. Process. Lett..

[43]  Jeffrey D Ullma.  Computational Aspects of VLSI , 1984 .

[44]  Cem Evrendilek,et al.  Vertex Separators for Partitioning a Graph , 2008, Sensors.

[45]  S.D. Muller,et al.  Step size adaptation in evolution strategies using reinforcement learning , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).

[46]  Mauro Brunato,et al.  Reactive Search and Intelligent Optimization , 2008 .