Workshop report on basic research needs for scientific machine learning: Core technologies for artificial intelligence

N Baker, F Alexander, T Bremer, A Hagberg… - 2019 - osti.gov
Scientific Machine Learning (SciML) and Artificial Intelligence (AI) will have broad use and
transformative effects across the Department of Energy. Accordingly, the January 2018 Basic …

Minimal variance sampling with provable guarantees for fast training of graph neural networks

W Cong, R Forsati, M Kandemir… - Proceedings of the 26th …, 2020 - dl.acm.org
Sampling methods (eg, node-wise, layer-wise, or subgraph) has become an indispensable
strategy to speed up training large-scale Graph Neural Networks (GNNs). However, existing …

Learning nonlocal constitutive models with neural networks

XH Zhou, J Han, H Xiao - Computer Methods in Applied Mechanics and …, 2021 - Elsevier
Constitutive and closure models play important roles in computational mechanics and
computational physics in general. Classical constitutive models for solid and fluid materials …

Optimal subsampling with influence functions

D Ting, E Brochu - Advances in neural information …, 2018 - proceedings.neurips.cc
Subsampling is a common and often effective method to deal with the computational
challenges of large datasets. However, for most statistical models, there is no well-motivated …

Distributed learning systems with first-order methods

J Liu, C Zhang - Foundations and Trends® in Databases, 2020 - nowpublishers.com
Scalable and efficient distributed learning is one of the main driving forces behind the recent
rapid advancement of machine learning and artificial intelligence. One prominent feature of …

Safe adaptive importance sampling

SU Stich, A Raj, M Jaggi - Advances in Neural Information …, 2017 - proceedings.neurips.cc
Importance sampling has become an indispensable strategy to speed up optimization
algorithms for large-scale applications. Improved adaptive variants--using importance …

On random embeddings and their application to optimisation

Z Shao - arXiv preprint arXiv:2206.03371, 2022 - arxiv.org
Random embeddings project high-dimensional spaces to low-dimensional ones; they are
careful constructions which allow the approximate preservation of key properties, such as …

Conditional Mixture Path Guiding for Differentiable Rendering

Z Fan, P Shi, M Guo, R Fu, Y Guo, J Guo - ACM Transactions on …, 2024 - dl.acm.org
The efficiency of inverse optimization in physically based differentiable rendering heavily
depends on the variance of Monte Carlo estimation. Despite recent advancements …

On feedback sample selection methods allowing lightweight digital predistorter adaptation

J Kral, T Gotthans, R Marsalek… - … on Circuits and …, 2020 - ieeexplore.ieee.org
In modern communication systems advanced techniques such as digital predistortion (DPD)
are required to satisfy stringent demands on transmitter linearity and efficiency. DPD …

Sampling with replacement vs Poisson sampling: a comparative study in optimal subsampling

J Wang, J Zou, HY Wang - IEEE Transactions on Information …, 2022 - ieeexplore.ieee.org
Faced with massive data, subsampling is a commonly used technique to improve
computational efficiency, and using nonuniform subsampling probabilities is an effective …