Stochastic Pareto local search: Pareto neighbourhood exploration and perturbation strategies

Pareto local search (PLS) methods are local search algorithms for multi-objective combinatorial optimization problems based on the Pareto dominance criterion. PLS explores the Pareto neighbourhood of a set of non-dominated solutions until it reaches a local optimal Pareto front. In this paper, we discuss and analyse three different Pareto neighbourhood exploration strategies: best, first, and neutral improvement. Furthermore, we introduce a deactivation mechanism that restarts PLS from an archive of solutions rather than from a single solution in order to avoid the exploration of already explored regions. To escape from a local optimal solution set we apply stochastic perturbation strategies, leading to stochastic Pareto local search algorithms (SPLS). We consider two perturbation strategies: mutation and path-guided mutation. While the former is unbiased, the latter is biased towards preserving common substructures between 2 solutions. We apply SPLS on a set of large, correlated bi-objective quadratic assignment problems (bQAPs) and observe that SPLS significantly outperforms multi-start PLS. We investigate the reason of this performance gain by studying the fitness landscape structure of the bQAPs using random walks. The best performing method uses the stochastic perturbation algorithms, the first improvement Pareto neigborhood exploration and the deactivation technique.

[1]  Thomas Stützle,et al.  A study of stochastic local search algorithms for the biobjective QAP with correlated flow matrices , 2006, Eur. J. Oper. Res..

[2]  Johannes Bader,et al.  Hypervolume-based search for multiobjective optimization: Theory and methods , 2010 .

[3]  Dirk Thierens,et al.  Path-Guided Mutation for Stochastic Pareto Local Search Algorithms , 2010, PPSN.

[4]  Andrzej Jaszkiewicz,et al.  Pareto memetic algorithm with path relinking for bi-objective traveling salesperson problem , 2009, Eur. J. Oper. Res..

[5]  Marco Laumanns,et al.  On Sequential Online Archiving of Objective Vectors , 2011, EMO.

[6]  El-Ghazali Talbi,et al.  On dominance-based multiobjective local search: design, implementation and experimental analysis on scheduling and traveling salesman problems , 2012, J. Heuristics.

[7]  Thomas Stützle,et al.  A hybrid TP+PLS algorithm for bi-objective flow-shop scheduling problems , 2011, Comput. Oper. Res..

[8]  Evripidis Bampis,et al.  A Dynasearch Neighborhood for the Bicriteria Traveling Salesman Problem , 2004, Metaheuristics for Multiobjective Optimisation.

[9]  David W. Corne,et al.  Instance Generators and Test Suites for the Multiobjective Quadratic Assignment Problem , 2003, EMO.

[10]  Bernd Freisleben,et al.  Fitness landscape analysis and memetic algorithms for the quadratic assignment problem , 2000, IEEE Trans. Evol. Comput..

[11]  Thomas Stützle,et al.  Pareto Local Search Algorithms for Anytime Bi-objective Optimization , 2012, EvoCOP.

[12]  Pierre Hansen,et al.  First vs. best improvement: An empirical study , 1999, Discret. Appl. Math..

[13]  David E. Goldberg,et al.  Alleles, loci and the traveling salesman problem , 1985 .

[14]  Nicola Beume,et al.  On the Complexity of Computing the Hypervolume Indicator , 2009, IEEE Transactions on Evolutionary Computation.

[15]  David E. Goldberg,et al.  AllelesLociand the Traveling Salesman Problem , 1985, ICGA.

[16]  Horst W. Hamacher,et al.  On spanning tree problems with multiple objectives , 1994, Ann. Oper. Res..

[17]  David Corne,et al.  Bounded Pareto Archiving: Theory and Practice , 2004, Metaheuristics for Multiobjective Optimisation.

[18]  Thomas Stützle,et al.  Pareto Local Optimum Sets in the Biobjective Traveling Salesman Problem: An Experimental Study , 2004, Metaheuristics for Multiobjective Optimisation.

[19]  Kalyanmoy Deb,et al.  Evaluating the -Domination Based Multi-Objective Evolutionary Algorithm for a Quick Computation of Pareto-Optimal Solutions , 2005, Evolutionary Computation.

[20]  Nicola Beume,et al.  SMS-EMOA: Multiobjective selection based on dominated hypervolume , 2007, Eur. J. Oper. Res..

[21]  Eckart Zitzler,et al.  HypE: An Algorithm for Fast Hypervolume-Based Many-Objective Optimization , 2011, Evolutionary Computation.

[22]  Thomas Stützle,et al.  A review of metrics on permutations for search landscape analysis , 2007, Comput. Oper. Res..

[23]  Xavier Gandibleux,et al.  Metaheuristics for Multiobjective Optimisation , 2004, Lecture Notes in Economics and Mathematical Systems.

[24]  Peter J. Fleming,et al.  On the Performance Assessment and Comparison of Stochastic Multiobjective Optimizers , 1996, PPSN.

[25]  Thomas Stützle,et al.  On local optima in multiobjective combinatorial optimization problems , 2007, Ann. Oper. Res..

[26]  Thomas Stützle,et al.  On the Design of ACO for the Biobjective Quadratic Assignment Problem , 2004, ANTS Workshop.

[27]  Marco Laumanns,et al.  Performance assessment of multiobjective optimizers: an analysis and review , 2003, IEEE Trans. Evol. Comput..

[28]  Jacques Teghem,et al.  Two-phase Pareto local search for the biobjective traveling salesman problem , 2010, J. Heuristics.

[29]  Thomas Stützle,et al.  Hybrid Population-Based Algorithms for the Bi-Objective Quadratic Assignment Problem , 2006, J. Math. Model. Algorithms.

[30]  Dirk Thierens,et al.  Generalized adaptive pursuit algorithm for genetic pareto local search algorithms , 2011, GECCO '11.

[31]  Thomas St Pareto Local Search Algorithms for Anytime Bi-Objective Optimization , 2011 .

[32]  Matthieu Basseur Design of cooperative algorithms for multi-objective optimization: application to the flow-shop scheduling problem , 2006, 4OR.

[33]  Tobias Friedrich,et al.  An Efficient Algorithm for Computing Hypervolume Contributions , 2010, Evolutionary Computation.

[34]  Marco Laumanns,et al.  PISA: A Platform and Programming Language Independent Interface for Search Algorithms , 2003, EMO.

[35]  Kiyoshi Tanaka,et al.  Random Bit Climbers on Multiobjective MNK-Landscapes: Effects of Memory and Population Climbing , 2005, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[36]  Nicola Beume,et al.  An EMO Algorithm Using the Hypervolume Measure as Selection Criterion , 2005, EMO.

[37]  Abdullah Alsheddy,et al.  Guided Pareto Local Search and its Application to the 0/1 Multi-objective Knapsack Problems , 2009 .

[38]  P. W. Poon,et al.  Genetic algorithm crossover operators for ordering applications , 1995, Comput. Oper. Res..

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