Efficient covariance matrix update for variable metric evolution strategies

DOI10.1007/s10994-009-5102-1zbMath1470.68183OpenAlexW2140671219MaRDI QIDQ1009335

Thorsten Suttorp, Christian Igel, Nikolaus Hansen

Publication date: 31 March 2009

Published in: Machine Learning (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1007/s10994-009-5102-1

zbMATH Keywords

stochastic optimizationmulti-objective optimizationreinforcement learningevolutionary algorithmcovariance matrix adaptationevolution strategyvariable metric algorithm

Mathematics Subject Classification ID

Evolutionary algorithms, genetic algorithms (computational aspects) (68W50) Multi-objective and goal programming (90C29) Learning and adaptive systems in artificial intelligence (68T05) Approximation methods and heuristics in mathematical programming (90C59) Problem solving in the context of artificial intelligence (heuristics, search strategies, etc.) (68T20) Analysis of variance and covariance (ANOVA) (62J10)

Related Items

No Free Lunch Theorems: Limitations and Perspectives of Metaheuristics, Principled Design of Continuous Stochastic Search: From Theory to Practice, Auxiliary model-based iterative estimation algorithms for nonlinear systems using the covariance matrix adaptation strategy, Global linear convergence of evolution strategies with recombination on scaling-invariant functions, An efficient conjugate gradient based Cholesky CMA‐ES estimation algorithm for nonlinear systems, Toward Nonlinear Local Reinforcement Learning Rules Through Neuroevolution, Speeding up many-objective optimization by Monte Carlo approximations, Variable metric evolution strategies by mutation matrix adaptation

Uses Software

• CMA-ES
• SMS-EMOA
• SHARK

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• SMS-EMOA: multiobjective selection based on dominated hypervolume
• Lower bounds for randomized direct search with isotropic sampling
• On classes of functions for which no free lunch results hold
• Evolution strategies. A comprehensive introduction
• Learning probability distributions in continuous evolutionary algorithms -- a comparative review
• Convergence results for the (1,\(\lambda\))-SA-ES using the theory of \(\varphi\)-irreducible Markov chains
• Optimization of dynamic neural fields
• A comparison of evolution strategies with other direct search methods in the presence of noise
• Multi-objective machine learning
• Weighted multirecombination evolution strategies
• How the (1+1) ES using isotropic mutations minimizes positive definite quadratic forms
• Can the Measure of ∪ n 1 [ a i , b i be Computed in Less Than O(n logn) Steps?]
• Lower Bounds for Hit-and-Run Direct Search
• Updating the Inverse of a Matrix
• New Upper Bounds in Klee’s Measure Problem
• Evolutionary reinforcement learning of artificial neural networks
• Evolutionary Multi-Criterion Optimization
• Evolutionary Multi-Criterion Optimization
• A comparison of evolution strategies and backpropagation for neural network training