Single- and multiobjective evolutionary optimization assisted by Gaussian random field metamodels

By Michael T. M. Emmerich, Kyriakos C. Giannakoglou, Boris Naujoks. IEEE Transactions on Evolutionary Computation, 2006

783 citations recorded by the source corpus.

Subfields: Computational Theory and Mathematics, Mechanical Engineering, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/55e46bfc236f14cce710931af789d7fd0ad8e2fb

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Mechanical Engineering

In Statistics and Probability

Citations in the corpus

A tutorial on multiobjective optimization: fundamentals and evolutionary methods (2018, background): “Multiobjective evolutionary algorithms (MOEAs) generalize this idea, and typically they are designed to gradually approach sets of Pareto optimal solutions that are well-distributed across the Pareto front.”
Multiobjective Optimization on a Limited Budget of Evaluations Using Model-Assisted -Metric Selection (2008, methodology): “[10,11,12] use local DACE models to determine lower and upper confidence bounds for a prescreening of solutions within classical multiobjective evolutionary algorithms (MOEA, EMOA), such as NSGA-II, ε-MOEA, and SMS-EMOA.”
Surrogate-assisted evolutionary computation: Recent advances and future challenges (2011, positive): “This idea of using coarse-to-fine surrogates has been introduced into a surrogateassisted evolutionary search in [42,62], where surrogates are used for evolutionary multi-objective optimization.”
CENTRAL FORCE OPTIMIZATION: A NEW META-HEURISTIC WITH APPLICATIONS IN APPLIED ELECTROMAGNETICS (2007, neutral): “The modified Keane’s Bump appears [15], and the Colville Function in [16].”
A Random Forest-Assisted Evolutionary Algorithm for Data-Driven Constrained Multiobjective Combinatorial Optimization of Trauma Systems (2020, methodology): “based model management strategies, promising and uncertain solutions according to the surrogate model [26]–[28] are evaluated using the expensive objective functions.”
Expensive Multiobjective Optimization by MOEA/D With Gaussian Process Model (2010, methodology): “Keane [21] and Emmerich et al [23] have generalized the probability of improvement and the expected improvement to multiobjective optimization.”
Multi-objective Optimisation Using Evolutionary Algorithms: An Introduction (2011, methodology): “Other successful meta-modeling implementations for multi-objective optimization based on Kriging and response surface methodologies exist [70, 71].”
A Gaussian Process Surrogate Model Assisted Evolutionary Algorithm for Medium Scale Expensive Optimization Problems (2014, methodology): “, [1], [8], [10], [11]), generates new points mainly for optimization purposes but not for modeling.”
Evolutionary Large-Scale Multi-Objective Optimization: A Survey (2021, methodology): “This is mainly due to the adopted surrogate models (e.g. Kriging [43]), which are inefficient to learn the mapping between a large number of decision variables and multiple objective functions.”
Resampling Methods for Meta-Model Validation with Recommendations for Evolutionary Computation (2012, methodology): “Additionally, uncertainty estimates of Kriging meta-model predictions can be used for prescreening promising candidate solutions for single objective and multi-objective EAs (Emmerich et al., 2006).”
A benchmark of kriging-based infill criteria for noisy optimization (2013, methodology): “It also has shown successful applications in kriging-based multiobjective optimization [8, 27].”
Latency-Aware Neural Architecture Search with Multi-Objective Bayesian Optimization (2021, background): “…functions include the expected improvement (EI) with respect to a scalarized objective (Knowles, 2006), expected hypervolume improvement (EHVI) (Emmerich et al., 2006; Daulton et al., 2020), and information gain with respect to the Pareto frontier (Hernandez-Lobato et al., 2016; Suzuki et al.,…”
Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization (2020, background): “Since ∆( { f ( x ) } , P , r ) is often a non-rectangular polytope, HVI is typically computed by partitioning the non-dominated space into disjoint axis-parallel rectangles (see Figure 1a) and use piece-wise integration [17].”
A Surrogate-Assisted Reference Vector Guided Evolutionary Algorithm for Computationally Expensive Many-Objective Optimization (2018, methodology): “If the Kriging model is used, one can select solutions that maximize the expected improvement [47], the probability of improvement [12], and hypervolume improvement [18].”
Recent Advances in Bayesian Optimization (2022, methodology): “In [42], Monte Carlo integration is adopted to approximate the EHVI.”