An Empirical Study of Off-Line Configuration and On-Line Adaptation in Operator Selection

Automating the process of finding good parameter settings is important in the design of high-performing algorithms. These automatic processes can generally be categorized into off-line and on-line methods. Off-line configuration consists in learning and selecting the best setting in a training phase, and usually fixes it while solving an instance. On-line adaptation methods on the contrary vary the parameter setting adaptively during each algorithm run. In this work, we provide an empirical study of both approaches on the operator selection problem, explore the possibility of varying parameter value by a non-adaptive distribution tuned off-line, and incorporate the off-line with on-line approaches. In particular, using an off-line tuned distribution to vary parameter values at runtime appears to be a promising idea for automatic configuration.

[1]  Martin J. Oates,et al.  On Fitness Distributions and Expected Fitness Gain of Mutation Rates in Parallel Evolutionary Algorithms , 2002, PPSN.

[2]  Carlos Ansótegui,et al.  A Gender-Based Genetic Algorithm for the Automatic Configuration of Algorithms , 2009, CP.

[3]  Thomas Stützle,et al.  An analysis of parameter adaptation in reactive tabu search , 2014, Int. Trans. Oper. Res..

[4]  Kevin Leyton-Brown,et al.  SATzilla: Portfolio-based Algorithm Selection for SAT , 2008, J. Artif. Intell. Res..

[5]  Jens Gottlieb,et al.  Evolutionary Computation in Combinatorial Optimization , 2006, Lecture Notes in Computer Science.

[6]  Peter J. Angeline,et al.  Adaptive and Self-adaptive Evolutionary Computations , 1995 .

[7]  Roberto Rossi,et al.  Synthesizing Filtering Algorithms for Global Chance-Constraints , 2009, CP.

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

[9]  Michèle Sebag,et al.  Extreme Value Based Adaptive Operator Selection , 2008, PPSN.

[10]  Zhi Yuan,et al.  Reinforcement learning for adaptive operator selection in memetic search applied to quadratic assignment problem , 2014, GECCO.

[11]  Zhi Yuan,et al.  Automated Parameter Tuning Framework for Heterogeneous and Large Instances: Case Study in Quadratic Assignment Problem , 2013, LION.

[12]  Simon M. Lucas,et al.  Parallel Problem Solving from Nature - PPSN X, 10th International Conference Dortmund, Germany, September 13-17, 2008, Proceedings , 2008, PPSN.

[13]  Dirk Thierens,et al.  An Adaptive Pursuit Strategy for Allocating Operator Probabilities , 2005, BNAIC.

[14]  Catherine C. McGeoch.  Analyzing algorithms by simulation: variance reduction techniques and simulation speedups , 1992, CSUR.

[15]  Zbigniew Michalewicz,et al.  Parameter Control in Evolutionary Algorithms , 2007, Parameter Setting in Evolutionary Algorithms.

[16]  Michèle Sebag,et al.  Toward comparison-based adaptive operator selection , 2010, GECCO '10.

[17]  Thomas Bartz-Beielstein,et al.  Experimental Methods for the Analysis of Optimization Algorithms , 2010 .

[18]  Vincenzo Cutello,et al.  Parallel Problem Solving from Nature - PPSN XII , 2012, Lecture Notes in Computer Science.

[19]  Michèle Sebag,et al.  Dynamic Multi-Armed Bandits and Extreme Value-Based Rewards for Adaptive Operator Selection in Evolutionary Algorithms , 2009, LION.

[20]  Zbigniew Michalewicz,et al.  Parameter Setting in Evolutionary Algorithms , 2007, Studies in Computational Intelligence.

[21]  Thomas Stützle,et al.  F-Race and Iterated F-Race: An Overview , 2010, Experimental Methods for the Analysis of Optimization Algorithms.

[22]  Martin L. Puterman,et al.  Markov Decision Processes: Discrete Stochastic Dynamic Programming , 1994 .

[23]  Roberto Battiti,et al.  The Reactive Tabu Search , 1994, INFORMS J. Comput..

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

[25]  Thomas Stützle,et al.  Continuous optimization algorithms for tuning real and integer parameters of swarm intelligence algorithms , 2011, Swarm Intelligence.

[26]  Thomas Stützle,et al.  New Benchmark Instances for the QAP and the Experimental Analysis of Algorithms , 2004, EvoCOP.

[27]  Éric D. Taillard,et al.  Robust taboo search for the quadratic assignment problem , 1991, Parallel Comput..

[28]  Panos M. Pardalos,et al.  Quadratic Assignment and Related Problems , 1994 .

[29]  Thomas Stützle,et al.  A Racing Algorithm for Configuring Metaheuristics , 2002, GECCO.

[30]  Holger H. Hoos,et al.  Programming by optimization , 2012, Commun. ACM.

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

[32]  Mauro Birattari,et al.  Tuning Metaheuristics - A Machine Learning Perspective , 2009, Studies in Computational Intelligence.

[33]  Juan Julián Merelo Guervós,et al.  Parallel Problem Solving from Nature — PPSN VII , 2002, Lecture Notes in Computer Science.

[34]  Marimuthu Palaniswami,et al.  Computational Intelligence: A Dynamic System Perspective , 1995 .

[35]  Franz Rendl,et al.  QAPLIB – A Quadratic Assignment Problem Library , 1997, J. Glob. Optim..

[36]  Thomas Bartz-Beielstein,et al.  Sequential Model-Based Parameter Optimization: an Experimental Investigation of Automated and Interactive Approaches , 2010, Experimental Methods for the Analysis of Optimization Algorithms.

[37]  Thomas Stützle,et al.  A critical analysis of parameter adaptation in ant colony optimization , 2011, Swarm Intelligence.

[38]  Thomas Stützle,et al.  Off-line and On-line Tuning: A Study on Operator Selection for a Memetic Algorithm Applied to the QAP , 2011, EvoCOP.

[39]  Thomas Stützle,et al.  An analysis of post-selection in automatic configuration , 2013, GECCO '13.

[40]  Eric Monfroy,et al.  Autonomous Search , 2012, Springer Berlin Heidelberg.

[41]  Helio J. C. Barbosa,et al.  Adaptive Operator Selection at the Hyper-level , 2012, PPSN.