A New Infill Criterion Based on Sorting of Approximation Uncertainty for Expensive Evolutionary Many-objective Optimization.
Xiaotong LiuChaoli SunHao WangPublished in: CEC (2023)
Keyphrases
- error bounds
- inherent uncertainty
- closed form
- genetic algorithm
- multiple objectives
- high cost
- approximation algorithms
- marginal likelihood
- error prone
- approximation error
- optimization criterion
- posterior marginals
- feature selection
- computationally expensive
- uncertain data
- multiple criteria
- uncertain information
- probability measure
- data sets