Surrogate Model Training

Surrogate model training aims to create computationally efficient approximations of complex simulations or machine learning models, accelerating tasks like hyperparameter optimization and causal effect estimation. Current research focuses on improving surrogate accuracy and transferability, employing diverse architectures including neural networks (e.g., transformers, graph neural networks, and variational autoencoders) and exploring techniques like ensemble methods and latent representation learning to enhance performance. These advancements significantly impact various fields by enabling faster exploration of high-dimensional parameter spaces, facilitating more efficient model development, and allowing for more robust analysis of complex systems.

Papers

October 19, 2024

Testing the Efficacy of Hyperparameter Optimization Algorithms in Short-Term Load Forecasting
Tugrul Cabir Hakyemez, Omer Adar
Bayesian Optimization Hyperparameter Optimization Level Test Forecasting Performance Short Term Load Forecasting Surrogate Model Training

July 15, 2023

Why Does Little Robustness Help? Understanding and Improving Adversarial Transferability from Surrogate Training
Yechao Zhang, Shengshan Hu, Leo Yu Zhang, Junyu Shi, Minghui Li, Xiaogeng Liu, Wei Wan, Hai Jin
Adversarial Example Adversarial Training Adversarial Sample Robustness Problem Surrogate Model Training

March 27, 2023

Deep Ranking Ensembles for Hyperparameter Optimization
Abdus Salam Khazi, Sebastian Pineda Arango, Josif Grabocka
Hyperparameter Optimization Related Hyperparameters Hyperparameter Configuration Neural Network Surrogate Surrogate Model Training

August 9, 2022

Long-term Causal Effects Estimation via Latent Surrogates Representation Learning
Ruichu Cai, Weilin Chen, Zeqin Yang, Shu Wan, Chen Zheng, Xiaoqing Yang, Jiecheng Guo
Surrogate Model Long Term Causal Effect Causal Structure Surrogate Modeling Surrogate Model Training

July 25, 2022

VDL-Surrogate: A View-Dependent Latent-based Model for Parameter Space Exploration of Ensemble Simulations
Neng Shi, Jiayi Xu, Haoyu Li, Hanqi Guo, Jonathan Woodring, Han-Wei Shen
Surrogate Model Parameter Space Exploration Simulation Ensemble Surrogate Model Training

March 17, 2022

Attribute Surrogates Learning and Spectral Tokens Pooling in Transformers for Few-shot Learning
Yangji He, Weihan Liang, Dongyang Zhao, Hong-Yu Zhou, Weifeng Ge, Yizhou Yu, Wenqiang Zhang
Convolutional Neural Network Vision Transformer Transformer Megatron Decepticons Surrogate Modeling Visual Recognition Spectral Token Cascaded Transformer Surrogate Model Training

February 5, 2022

A Graph Neural Network Framework for Grid-Based Simulation
Haoyu Tang, Wennan Long
Graph Neural Network Model Grid Based Reservoir Simulation Well Control Surrogate Model Training

Surrogate Model Training

Papers

Testing the Efficacy of Hyperparameter Optimization Algorithms in Short-Term Load Forecasting

Why Does Little Robustness Help? Understanding and Improving Adversarial Transferability from Surrogate Training

Deep Ranking Ensembles for Hyperparameter Optimization

Long-term Causal Effects Estimation via Latent Surrogates Representation Learning

VDL-Surrogate: A View-Dependent Latent-based Model for Parameter Space Exploration of Ensemble Simulations

Attribute Surrogates Learning and Spectral Tokens Pooling in Transformers for Few-shot Learning

A Graph Neural Network Framework for Grid-Based Simulation