Surrogate Neural Network

Surrogate neural networks are artificial neural networks trained to approximate computationally expensive simulations or models, accelerating analyses and predictions across diverse scientific and engineering domains. Current research emphasizes improving surrogate accuracy and efficiency through techniques like active learning for optimized data selection, and accelerated algorithms such as those replacing Monte Carlo sampling with surrogate networks themselves. These advancements are crucial for tackling complex problems in fields ranging from aircraft design and ocean modeling to hypersonic vehicle trajectory planning, enabling faster, more cost-effective simulations and optimization processes.

Papers

July 10, 2024

Feasibility Study on Active Learning of Smart Surrogates for Scientific Simulations
Pradeep Bajracharya, Javier Quetzalcóatl Toledo-Marín, Geoffrey Fox, Shantenu Jha, Linwei Wang
Deep Neural Network Active Learning Surrogate Model Feasibility Study Surrogate Neural Network Surrogate Study Assembling

February 12, 2024

Accelerated Smoothing: A Scalable Approach to Randomized Smoothing
Devansh Bhardwaj, Kshitiz Kaushik, Sarthak Gupta
Adversarial Attack Randomized Smoothing Scalable Approach Smoothed Classifier Robustness Certificate Smoothing Algorithm Surrogate Neural Network

January 11, 2024

Surrogate Neural Networks Local Stability for Aircraft Predictive Maintenance
Mélanie Ducoffe, Guillaume Povéda, Audrey Galametz, Ryma Boumazouza, Marion-Cécile Martin, Julien Baris, Derk Daverschot, Eugene O'Higgins
Native Robustness Surrogate Model Neural Network Verification Efficient Verification Surrogate Neural Network

November 10, 2023

Surrogate Neural Networks to Estimate Parametric Sensitivity of Ocean Models
Yixuan Sun, Elizabeth Cucuzzella, Steven Brus, Sri Hari Krishna Narayanan, Balu Nadiga, Luke Van Roekel, Jan Hückelheim, Sandeep Madireddy
Ocean Dynamic Model Sensitivity Neural Surrogate Ocean Modeling Surrogate Neural Network

March 30, 2023

Surrogate Neural Networks for Efficient Simulation-based Trajectory Planning Optimization
Evelyn Ruff, Rebecca Russell, Matthew Stoeckle, Piero Miotto, Jonathan P. How
High Efficiency Surrogate Model Trajectory Planning Efficient Trajectory Simulation Based Optimization Minimum Time Trajectory Surrogate Neural Network

September 7, 2022

SmOOD: Smoothness-based Out-of-Distribution Detection Approach for Surrogate Neural Networks in Aircraft Design
Houssem Ben Braiek, Ali Tfaily, Foutse Khomh, Thomas Reid, Ciro Guida
Distribution Detection Surrogate Model Aircraft Design Aerodynamic Shape Optimization Surrogate Neural Network

June 28, 2022

How to Steer Your Adversary: Targeted and Efficient Model Stealing Defenses with Gradient Redirection
Mantas Mazeika, Bo Li, David Forsyth
Adversary Agent Model Stealing Attack Surrogate Neural Network Computational Overhead Prior Defense

June 9, 2022

GSmooth: Certified Robustness against Semantic Transformations via Generalized Randomized Smoothing
Zhongkai Hao, Chengyang Ying, Yinpeng Dong, Hang Su, Jun Zhu, Jian Song
Native Robustness Randomized Smoothing Robustness Certification Preserving Transformation Perturbation Augmentation Surrogate Neural Network

March 16, 2022

Attacking deep networks with surrogate-based adversarial black-box methods is easy
Nicholas A. Lord, Romain Mueller, Luca Bertinetto
Adversarial Attack Deep Network Surrogate Model Transfer Attack Surrogate Neural Network

Surrogate Neural Network

Papers

Feasibility Study on Active Learning of Smart Surrogates for Scientific Simulations

Accelerated Smoothing: A Scalable Approach to Randomized Smoothing

Surrogate Neural Networks Local Stability for Aircraft Predictive Maintenance

Surrogate Neural Networks to Estimate Parametric Sensitivity of Ocean Models

Surrogate Neural Networks for Efficient Simulation-based Trajectory Planning Optimization

SmOOD: Smoothness-based Out-of-Distribution Detection Approach for Surrogate Neural Networks in Aircraft Design

How to Steer Your Adversary: Targeted and Efficient Model Stealing Defenses with Gradient Redirection

GSmooth: Certified Robustness against Semantic Transformations via Generalized Randomized Smoothing

Attacking deep networks with surrogate-based adversarial black-box methods is easy