Multiobjective Optimization on a Budget of 250 Evaluations

In engineering and other ‘real-world' applications, multiobjective optimization problems must frequently be tackled on a tight evaluation budget — tens or hundreds of function evaluations, rather than thousands. In this paper, we investigate two algorithms that use advanced initialization and search strategies to operate better under these conditions. The first algorithm, Bin_MSOPS, uses a binary search tree to divide up the decision space, and tries to sample from the largest empty regions near ‘fit' solutions. The second algorithm, ParEGO, begins with solutions in a latin hypercube and updates a Gaussian processes surrogate model of the search landscape after every function evaluation, which it uses to estimate the solution of largest expected improvement. The two algorithms are tested using a benchmark suite of nine functions of two and three objectives — on a budget of only 250 function evaluations each, in total. Results indicate that the two algorithms search the space in very different ways and this can be used to understand performance differences. Both algorithms perform well but ParEGO comes out on top in seven of the nine test cases after 100 function evaluations, and on six after the first 250 evaluations.

[1]  N. Zheng,et al.  Global Optimization of Stochastic Black-Box Systems via Sequential Kriging Meta-Models , 2006, J. Glob. Optim..

[2]  Hisham M. Haddad,et al.  Proceedings of the 1999 ACM Symposium on Applied Computing, SAC'99, San Antonio, Texas, USA, February 28 - March 2, 1999 , 1999, ACM Symposium on Applied Computing.

[3]  Enrique Dunn,et al.  Multi-objective Sensor Planning for Efficient and Accurate Object Reconstruction , 2004, EvoWorkshops.

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

[5]  Ralph E. Steuer,et al.  An interactive weighted Tchebycheff procedure for multiple objective programming , 1983, Math. Program..

[6]  David E. Goldberg,et al.  Fitness Inheritance In Multi-objective Optimization , 2002, GECCO.

[7]  Marco Laumanns,et al.  Scalable Test Problems for Evolutionary Multiobjective Optimization , 2005, Evolutionary Multiobjective Optimization.

[8]  Henry P. Wynn,et al.  [Design and Analysis of Computer Experiments]: Rejoinder , 1989 .

[9]  Kaisa Miettinen,et al.  Nonlinear multiobjective optimization , 1998, International series in operations research and management science.

[10]  D. Kell,et al.  Explanatory optimization of protein mass spectrometry via genetic search. , 2003, Analytical chemistry.

[11]  Donald R. Jones,et al.  Efficient Global Optimization of Expensive Black-Box Functions , 1998, J. Glob. Optim..

[12]  António Gaspar-Cunha,et al.  A Hybrid Multi-Objective Evolutionary Algorithm Using an Inverse Neural Network , 2004, Hybrid Metaheuristics.

[13]  Kok Wai Wong,et al.  Surrogate-Assisted Evolutionary Optimization Frameworks for High-Fidelity Engineering Design Problems , 2005 .

[14]  Bernard De Baets,et al.  Is Fitness Inheritance Useful for Real-World Applications? , 2003, EMO.

[15]  Michael Emmerich,et al.  Metamodel Assisted Multiobjective Optimisation Strategies and their Application in Airfoil Design , 2004 .

[16]  Bernhard Sendhoff,et al.  On Test Functions for Evolutionary Multi-objective Optimization , 2004, PPSN.

[17]  M. Hansen,et al.  Evaluating the quality of approximations to the non-dominated set , 1998 .

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

[19]  Joshua D. Knowles,et al.  ParEGO: a hybrid algorithm with on-line landscape approximation for expensive multiobjective optimization problems , 2006, IEEE Transactions on Evolutionary Computation.

[20]  William H. Press,et al.  Numerical recipes in C. The art of scientific computing , 1987 .

[21]  António Gaspar-Cunha,et al.  A Multi-Objective Evolutionary Algorithm Using Neural Networks to Approximate Fitness Evaluations , 2005, Int. J. Comput. Syst. Signals.

[22]  Yaochu Jin,et al.  A comprehensive survey of fitness approximation in evolutionary computation , 2005, Soft Comput..

[23]  Kalyanmoy Deb,et al.  Self-Adaptive Genetic Algorithms with Simulated Binary Crossover , 2001, Evolutionary Computation.

[24]  Masahiro Fujita,et al.  Autonomous evolution of dynamic gaits with two quadruped robots , 2005, IEEE Transactions on Robotics.

[25]  Gary B. Lamont,et al.  Multiobjective evolutionary algorithm test suites , 1999, SAC '99.

[26]  Dirk Thierens,et al.  Multi-objective Optimization with the Naive MIDEA , 2006 .

[27]  R. J. Gilbert,et al.  Efficient Improvement of Silage Additives by Using Genetic Algorithms , 2000, Applied and Environmental Microbiology.

[28]  Hans-Paul Schwefel,et al.  Parallel Problem Solving from Nature — PPSN IV , 1996, Lecture Notes in Computer Science.

[29]  H. Beyer.  Evolutionary algorithms in noisy environments : theoretical issues and guidelines for practice , 2000 .

[30]  Yaochu Jin,et al.  Knowledge incorporation in evolutionary computation , 2005 .

[31]  P. Coveney,et al.  Combinatorial searches of inorganic materials using the ink-jet printer: science, philosophy and technology , 2001 .

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

[33]  C. Wandrey,et al.  Medium Optimization by Genetic Algorithm for Continuous Production of Formate Dehydrogenase , 1995 .

[34]  Andrew W. Moore,et al.  A Nonparametric Approach to Noisy and Costly Optimization , 2000, ICML.

[35]  Petros Koumoutsakos,et al.  Self-organizing Maps for Pareto Optimization of Airfoils , 2002, PPSN.

[36]  Dirk Thierens,et al.  Multi-objective Optimization with the Naive $$ \mathbb{M} $$ ID $$ \mathbb{E} $$ A , 2006, Towards a New Evolutionary Computation.

[37]  Joshua D. Knowles,et al.  Closed-loop, multiobjective optimization of analytical instrumentation: gas chromatography/time-of-flight mass spectrometry of the metabolomes of human serum and of yeast fermentations. , 2005, Analytical chemistry.

[38]  Bernhard Sendhoff,et al.  A framework for evolutionary optimization with approximate fitness functions , 2002, IEEE Trans. Evol. Comput..

[39]  Evan J. Hughes.  Multi-objective Binary Search Optimisation , 2003, EMO.

[40]  Shigeru Obayashi,et al.  Optimization of Combustion Chamber for Diesel Engine Using Kriging Model , 2006 .

[41]  E. Hughes.  Multiple single objective Pareto sampling , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[42]  Marco Laumanns,et al.  Bayesian Optimization Algorithms for Multi-objective Optimization , 2002, PPSN.

[43]  Martin J. Oates,et al.  PESA-II: region-based selection in evolutionary multiobjective optimization , 2001 .

[44]  William H. Press,et al.  The Art of Scientific Computing Second Edition , 1998 .

[45]  Joshua D. Knowles,et al.  On metrics for comparing nondominated sets , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).

[46]  William H. Press,et al.  Book-Review - Numerical Recipes in Pascal - the Art of Scientific Computing , 1989 .

[47]  Kalyanmoy Deb,et al.  An Investigation of Niche and Species Formation in Genetic Function Optimization , 1989, ICGA.

[48]  I. C. Parmee.  Adaptive Computing in Design and Manufacture , 1998 .

[49]  Douglas C. Montgomery,et al.  Response Surface Methodology: Process and Product Optimization Using Designed Experiments , 1995 .

[50]  Joshua D. Knowles,et al.  Closed-loop, multiobjective optimization of two-dimensional gas chromatography/mass spectrometry for serum metabolomics. , 2007, Analytical chemistry.