On the Design of a Master-Worker Adaptive Algorithm Selection Framework

We investigate the design of a master-worker schemes for adaptive algorithm selection with the following two-fold goal: (i) choose accurately from a given portfolio a set of operators to be executed in parallel, and consequently (ii) take full advantage of the compute power offered by the underlying distributed environment. In fact, it is still an open issue to design online distributed strategies that are able to optimally assign operators to parallel compute resources when distributively solving a given optimization problem. In our proposed framework, we adopt a reward-based perspective and investigate at what extent the average or maximum rewards collected at the master from the workers are appropriate. Moreover, we investigate the design of both homogeneous and heterogeneous scheme. Our comprehensive experimental study, conducted through a simulation-based methodology and using a recently proposed benchmark family for adaptive algorithm selection, reveals the accuracy of the proposed framework while providing new insights on the performance of distributed adaptive optimization algorithms.

[1]  Christopher Jankee,et al.  A Fitness Cloud Model for Adaptive Metaheuristic Selection Methods , 2016, PPSN.

[2]  Zbigniew Michalewicz,et al.  Evolutionary Algorithms in Engineering Applications , 1997, Springer Berlin Heidelberg.

[3]  Petr Posík,et al.  Online Black-Box Algorithm Portfolios for Continuous Optimization , 2014, PPSN.

[4]  Alex S. Fukunaga,et al.  Evaluation of a randomized parameter setting strategy for island-model evolutionary algorithms , 2013, 2013 IEEE Congress on Evolutionary Computation.

[5]  Frédéric Saubion,et al.  Non stationary operator selection with island models , 2013, GECCO '13.

[6]  Michèle Sebag,et al.  Extreme compass and Dynamic Multi-Armed Bandits for Adaptive Operator Selection , 2009, 2009 IEEE Congress on Evolutionary Computation.

[7]  Michèle Sebag,et al.  Adaptive operator selection with dynamic multi-armed bandits , 2008, GECCO '08.

[8]  Michèle Sebag,et al.  Analyzing bandit-based adaptive operator selection mechanisms , 2010, Annals of Mathematics and Artificial Intelligence.

[9]  John J. Grefenstette,et al.  Optimization of Control Parameters for Genetic Algorithms , 1986, IEEE Transactions on Systems, Man, and Cybernetics.

[10]  Saúl Zapotecas Martínez,et al.  Traffic Signal Optimization: Minimizing Travel Time and Fuel Consumption , 2015, Artificial Evolution.

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

[12]  Jean-Michel Do,et al.  A Multi-Physics PWR Model for the Load Following , 2016 .

[13]  Günter Rudolph,et al.  Comparing Asynchronous and Synchronous Parallelization of the SMS-EMOA , 2016, PPSN.

[14]  Juan Julián Merelo Guervós,et al.  Randomized Parameter Settings for Heterogeneous Workers in a Pool-Based Evolutionary Algorithm , 2014, PPSN.

[15]  Frédéric Saubion,et al.  Simulating non-stationary operators in search algorithms , 2016, Appl. Soft Comput..

[16]  Dirk Sudholt,et al.  Parallel Evolutionary Algorithms , 2015, Handbook of Computational Intelligence.

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

[18]  Sébastien Vérel,et al.  DAMS: distributed adaptive metaheuristic selection , 2011, GECCO '11.

[19]  Lars Kotthoff,et al.  Algorithm Selection for Combinatorial Search Problems: A Survey , 2012, AI Mag..

[20]  Christopher Jankee,et al.  Distributed Adaptive Metaheuristic Selection: Comparisons of Selection Strategies , 2015, Artificial Evolution.

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

[22]  Marc Schoenauer,et al.  Asynchronous master/slave moeas and heterogeneous evaluation costs , 2012, GECCO '12.

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

[24]  Marc Parizeau,et al.  Analysis of a master-slave architecture for distributed evolutionary computations , 2006, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).