Jiaming Song
Willie Neiswanger
Jiashuo Liu
Description: D2C is a VAE-based generative model suitable for few-shot conditional generation.
diffusion (239) vae (132) generative modeling (5) contrastive representation learning (2)
In Bayesian Optimization (BO), the typical acquisition function requires a probabilistic surrogate model (such as Gaussian Processes (GP) ). In Likelihood-free Bayesian Optimization (LFBO) , the surrogate model is a deterministic model that directly reflects the acquisition function; this approach bypasses expensive GP inference yet results in similar sequential queries.
The acquisition function, a critical component in Bayesian optimization (BO), can often be written as the expectation of a utility function under a surrogate model. However, to ensure that acquisition functions are tractable to optimize, restrictions must be placed on the surrogate model and utility function. To extend BO to a broader class of models and utilities, we propose likelihood-free BO (LFBO), an approach based on likelihood-free inference.
LFBO directly models the acquisition function without having to separately perform inference with a probabilistic surrogate model. We show that computing the acquisition function in LFBO can be reduced to optimizing a weighted classification problem, where the weights correspond to the utility being chosen. LFBO outperforms various state-of-the-art black-box optimization methods on several real-world optimization problems. LFBO can also effectively leverage composite structures of the objective function, wh
Jiaming Song
Willie Neiswanger
Jiashuo Liu