RoMA: Robust Model Adaptation for Offline Model-based Optimization
arXiv:2110.14188
Abstract
We consider the problem of searching an input maximizing a black-box objective function given a static dataset of input-output queries. A popular approach to solving this problem is maintaining a proxy model, e.g., a deep neural network (DNN), that approximates the true objective function. Here, the main challenge is how to avoid adversarially optimized inputs during the search, i.e., the inputs where the DNN highly overestimates the true objective function. To handle the issue, we propose a new framework, coined robust model adaptation (RoMA), based on gradient-based optimization of inputs over the DNN. Specifically, it consists of two steps: (a) a pre-training strategy to robustly train the proxy model and (b) a novel adaptation procedure of the proxy model to have robust estimates for a specific set of candidate solutions. At a high level, our scheme utilizes the local smoothness prior to overcome the brittleness of the DNN. Experiments under various tasks show the effectiveness of RoMA compared with previous methods, obtaining state-of-the-art results, e.g., RoMA outperforms all at 4 out of 6 tasks and achieves runner-up results at the remaining tasks.
arXiv commentNeurIPS 2021
References in corpus (15)
- Auto-Encoding Variational Bayes
- Decoupled Weight Decay Regularization
- Conditional Generative Adversarial Nets
- Practical Bayesian Optimization of Machine Learning Algorithms
- Soft Actor-Critic Algorithms and Applications
- Certified Adversarial Robustness via Randomized Smoothing
- Scalable Bayesian Optimization Using Deep Neural Networks
- Tent: Fully Test-time Adaptation by Entropy Minimization
- Robust Large Margin Deep Neural Networks
- Spectral Norm Regularization for Improving the Generalizability of Deep Learning
- Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness
- Smooth Adversarial Training
- Improving Robustness Without Sacrificing Accuracy with Patch Gaussian Augmentation
- A case for new neural network smoothness constraints
- Offline Model-Based Optimization via Normalized Maximum Likelihood Estimation