Generative Surrogate

Generative surrogate models are computationally efficient alternatives to complex simulations, aiming to accurately reproduce the output of computationally expensive black-box functions or processes. Current research focuses on improving the fidelity of generated samples using architectures like normalizing flows and GANs, often incorporating convolutional layers or U-Net structures for higher-dimensional data. These models find applications in diverse fields, from optimizing challenging engineering designs (e.g., metasurfaces) and accelerating scientific discovery to enhancing adversarial attacks on machine learning models and improving active feature acquisition in resource-constrained settings. The ultimate goal is to create highly accurate and efficient surrogates that enable faster, cheaper, and more robust exploration of complex systems.

Papers

July 16, 2024

Global Optimisation of Black-Box Functions with Generative Models in the Wasserstein Space
Tigran Ramazyan, Mikhail Hushchyn, Denis Derkach
Generative Model Global Optimization Gradient Free Black Box Function Wasserstein Space Wasserstein Uncertainty Generative Surrogate

May 30, 2024

Convolutional L2LFlows: Generating Accurate Showers in Highly Granular Calorimeters Using Convolutional Normalizing Flows
Thorsten Buss, Frank Gaede, Gregor Kasieczka, Claudius Krause, David Shih
Normalizing Flow Convolutional Layer Fully Convolutional Autoregressive Flow Particle Shower Generative Surrogate

February 5, 2024

A Generative Approach to Surrogate-based Black-box Attacks
Raha Moraffah, Huan Liu
Adversarial Example Generative Approach Generative Surrogate

October 18, 2023

A Surrogate-Assisted Extended Generative Adversarial Network for Parameter Optimization in Free-Form Metasurface Design
Manna Dai, Yang Jiang, Feng Yang, Joyjit Chattoraj, Yingzhi Xia, Xinxing Xu, Weijiang Zhao, My Ha Dao, Yong Liu
Metasurface Design Parameter Optimization Generative Surrogate

February 27, 2023

Acquisition Conditioned Oracle for Nongreedy Active Feature Acquisition
Michael Valancius, Max Lennon, Junier Oliva
Sparse Reward Test Oracle Channel Acquisition Feature Acquisition Active Feature Acquisition Generative Surrogate

June 9, 2022

Deep Surrogate Assisted Generation of Environments
Varun Bhatt, Bryon Tjanaka, Matthew C. Fontaine, Stefanos Nikolaidis
Reinforcement Learning Environment Feature Quality Diversity Generative Surrogate Environment Generation