Bayesian Optimization over Discrete and Mixed Spaces via Probabilistic Reparameterization.
Samuel DaultonXingchen WanDavid ErikssonMaximilian BalandatMichael A. OsborneEytan BakshyPublished in: CoRR (2022)
Keyphrases
- posterior probability
- bayesian networks
- data driven
- optimization problems
- optimization process
- probabilistic model
- decision theory
- discrete valued
- bayesian estimation
- global optimization
- maximum likelihood
- maximum entropy
- optimization methods
- gaussian processes
- optimization algorithm
- continuous optimization
- efficient optimization
- uncertain data
- information theoretic
- random variables
- evolution strategy
- prior probabilities
- bi level
- continuous valued
- sequence prediction