Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization.
Samuel DaultonMaximilian BalandatEytan BakshyPublished in: NeurIPS (2020)
Keyphrases
- multi objective
- uniform design
- optimization algorithm
- nsga ii
- objective function
- evolutionary algorithm
- multiple objectives
- multiobjective optimization
- multi objective optimization
- evolutionary optimization
- optimization problems
- optimum design
- conflicting objectives
- genetic algorithm
- particle swarm
- particle swarm optimization
- engineering design problems
- multi objective evolutionary algorithms
- global optimization
- parallel processing
- multi objective problems
- multiobjective evolutionary algorithm
- pareto optimal
- efficient optimization
- optimization method
- estimation of distribution algorithms
- strength pareto evolutionary algorithm
- bayesian networks
- constrained optimization
- optimization process
- trade off
- maximum likelihood
- loss function
- multi objective evolutionary
- search algorithm
- significant improvement
- scheduling problem
- markov random field
- monte carlo sampling
- multi objective optimization problems
- shared memory
- metaheuristic
- differential evolution
- posterior probability
- distributed memory
- bi objective
- computer architecture
- gaussian processes
- parallel computing