Learning Large-scale Subsurface Simulations with a Hybrid Graph Network Simulator

Wu, Tong; Qinchen, Wang,; Zhang, Yinan; Ying, Rex; Cao, Kaidi; Sosič, Rok; Ridwan, Jalali,; Hamam, Hassan; Maučec, Marko; Leskovec, Jure

doi:10.1145/3534678.3539045

Cited by 9 publications

(6 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Methods: Plenty of studies have been devoted to learning to simulate complex physical systems using GNNs, including Interaction Network [14], NRI [133], HRN [186], DPI-Net [151], HOGN [213], GNS [214], C-GNS [212], HGNS [274], GNS* [2], and FIGNet [3]. However, all these methods adopt typical GNNs that are unaware of full symmetry in 3D word, and only a subset of them considers translation-equivariance.…”

Section: Physical Dynamics Simulationmentioning

confidence: 99%

Single-cell atlas of a non-human primate reveals new pathogenic mechanisms of COVID-19

Han

Wei

Liu

et al. 2020

Preprint

View full text Add to dashboard Cite

Stopping COVID-19 is a priority worldwide. Understanding which cell types are targeted by SARS-CoV-2 virus, whether interspecies differences exist, and how variations in cell state influence viral entry is fundamental for accelerating therapeutic and preventative approaches. In this endeavor, we profiled the transcriptome of nine tissues from a Macaca fascicularis monkey at single-cell resolution. The distribution of SARS-CoV-2 facilitators, ACE2 and TMRPSS2, in different cell subtypes showed substantial heterogeneity across lung, kidney, and liver. Through co-expression analysis, we identified immunomodulatory proteins such as IDO2 and ANPEP as potential SARS-CoV-2 targets responsible for immune cell exhaustion. Furthermore, single-cell chromatin accessibility analysis of the kidney unveiled a plausible link between IL6-mediated innate immune responses aiming to protect tissue and enhanced ACE2 expression that could promote viral entry. Our work constitutes a unique resource for understanding the physiology and pathophysiology of two phylogenetically close species, which might guide in the development of therapeutic approaches in humans.Bullet pointsWe generated a single-cell transcriptome atlas of 9 monkey tissues to study COVID-19.ACE2+TMPRSS2+ epithelial cells of lung, kidney and liver are targets for SARS-CoV-2.ACE2 correlation analysis shows IDO2 and ANPEP as potential therapeutic opportunities.We unveil a link between IL6, STAT transcription factors and boosted SARS-CoV-2 entry.

show abstract

Section: Physical Dynamics Simulationmentioning

confidence: 99%

Single-cell atlas of a non-human primate reveals new pathogenic mechanisms of COVID-19

Han

Wei

Liu

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Although this approach eliminates the distributional shift for rollouts of only 2 timesteps, the shift appears again for longer rollouts. Wu et al [2022c] address this by optimizing their Hybrid Graph Network Simulator (HGNS) using a multi-step objective in recurrently predicting several steps ahead. The total loss is a weighted sum of the loss at each time step, with the one-step loss weighted heaviest so that the optimization initially targets short-term predictions before fine-tuning for longer-term predictions.…”

Section: Existing Methods: Multiresolution Dynamicsmentioning

confidence: 99%

“…A second factor limiting applicability is that much of the literature focuses on problems with only one or two spatial dimensions despite the prevalence of three-dimensional problems in industrial applications of PDE modeling. While many architectures discussed in this section admit an immediate extension to three spatial dimensions, in practice, three-dimensional modeling presents obstacles in the form of limited memory that must be carefully handled [Wu et al 2022c;Lam et al 2022;. Beyond memory requirements, the optimization of a three-dimensional model is naturally more challenging than its two-dimensional counterpart due to the increased size of the search space.…”

Section: Existing Methods: Multiresolution Dynamicsmentioning

confidence: 99%

“…To address these shortcomings, deep learning models have emerged as a general framework to produce solutions orders of magnitude faster than their numerical counterparts. This efficiency is primarily achieved via the ability of neural networks to take significantly larger time steps during inference, learn on more coarse spatial discretizations compared to classical solvers Stachenfeld et al 2021], and use explicit forward methods instead of implicit methods Wu et al 2022c].…”

Section: Overviewmentioning

confidence: 99%

“…Multiscale dynamics arise in systems where physics evolve on a continuum from local to global scales [Stachenfeld et al 2021;Gupta and Brandstetter 2022;Li et al 2021b;Lam et al 2022;Pathak et al 2022;Rui Wang 2020;Li et al 2021c], while multiresolution dynamics occur in systems with fast-evolving, isolated regions that require greater resources for stable simulation ]. Solvers that use explicit schemes, where the state of the system at the next time step is computed from the current time step, encounter error in their inputs introduced by previous predictions and thus must consider methods for maintaining rollout stability [Sanchez-Gonzalez et al 2020;Brandstetter et al 2022c;Wu et al 2022c]. Equivariant architectures and training techniques enforce symmetries of the system, enabling improved generalization and sample complexity [Brandstetter et al 2022b;Ruhe et al 2023;Horie et al 2021;Helwig et al 2023].…”

Section: Overviewmentioning

confidence: 99%

See 2 more Smart Citations

Artificial intelligence in recommender systems

Zhang

Jin

2020

Complex Intell. Syst.

209

View full text Add to dashboard Cite

Recommender systems provide personalized service support to users by learning their previous behaviors and predicting their current preferences for particular products. Artificial intelligence (AI), particularly computational intelligence and machine learning methods and algorithms, has been naturally applied in the development of recommender systems to improve prediction accuracy and solve data sparsity and cold start problems. This position paper systematically discusses the basic methodologies and prevailing techniques in recommender systems and how AI can effectively improve the technological development and application of recommender systems. The paper not only reviews cutting-edge theoretical and practical contributions, but also identifies current research issues and indicates new research directions. It carefully surveys various issues related to recommender systems that use AI, and also reviews the improvements made to these systems through the use of such AI approaches as fuzzy techniques, transfer learning, genetic algorithms, evolutionary algorithms, neural networks and deep learning, and active learning. The observations in this paper will directly support researchers and professionals to better understand current developments and new directions in the field of recommender systems using AI.

show abstract