Results 161 to 170 of about 4,082,283 (306)

Graph neural networks and MSO

CoRR
We give an alternative proof for the existing result that recurrent graph neural networks working with reals have the same expressive power in restriction to monadic second-order logic MSO as the graded modal substitution calculus. The proof is based on constructing distributed automata that capture all MSO-definable node properties over trees. We also
Veeti Ahvonen, Damian Heiman, Antti Kuusisto   +2 more
openaire   +3 more sources

Toward Pore‐Engineered 3D‐Printed Materials for Sorption Water Harvesting, Interfacial Evaporation, and Radiative Cooling

Advanced Materials, EarlyView.
Water harvesting, radiative cooling, and interfacial solar evaporation are fundamentally governed by coupled heat, mass, and light transport processes. These processes are mediated by pore architecture, including pore size, connectivity, and hierarchical organization.
Dejan J. Trajkovski, Higgins M. Wilson, Abdallah Y. M. Ali, Jiaxing Xu, Ruzhu Wang, Primož Poredoš   +5 more
wiley   +1 more source

Graph Neural Networks for Bipartite Graphs

Bipartite graphs are a special type of graph data structure where vertices can be divided into two disjoint and independent sets, and each edge connects a vertex from one set to a vertex in the other set. They can be used to model many real-world applications such as user-item interaction networks, authorship networks, and product-customer networks ...
openaire   +2 more sources

Graph Neural Networks are Heuristics

CoRR
Graph neural networks are usually treated as auxiliaries for combinatorial optimization: they imitate algorithms, guide search, or supply scores to classical procedures. We show that this auxiliary role is not intrinsic. A GNN can itself be a heuristic.
Yimeng Min, Carla P. Gomes
openaire   +2 more sources

Utilizing correlation in space and time: Anomaly detection for Industrial Internet of Things (IIoT) via spatiotemporal gated graph attention network

Alexandria Engineering Journal
The Industrial Internet of Things (IIoT) infrastructure is inherently complex, often involving a multitude of sensors and devices. Ensuring the secure operation and maintenance of these systems is increasingly critical, making anomaly detection a vital ...
Yuxin Fan, Tingting Fu, Nikolai Izmailovich Listopad, Peng Liu, Sahil Garg, Mohammad Mehedi Hassan   +5 more
doaj   +1 more source

Bottom‐Up Engineering of a Human Neuromuscular System for Modeling Activity‐Induced Remodeling, Metabolic Stress, and Endothelial‐Modulated Excitability

Advanced Materials, EarlyView.
Using a bottom‐up approach, we engineered a human neuromuscular microchip to investigate distinct adaptive responses to neural stimulation and metabolic stress. Further integration of endothelial cells revealed their critical role in modulating neuromuscular excitability, establishing a comprehensive neurovascular‐muscular model.
Jinchul Ahn, Min‐seop Kim, Ju‐Hee Kim, Seok‐Hyeon Kang, Dain Lee, Hui‐Wen Liu, Seung‐cheol Shin, Kyuhwan Na, Yesl Jun, Oak‐Kee Hong, Ji Hun Yang, Jin‐A Kim, Hwanseok Jang, Hong Nam Kim, Roger D. Kamm, Ji‐Hoon Kim, Jang‐Won Son, Seok Chung   +17 more
wiley   +1 more source

Framework Flexibility on Water Sorption and Transport in ZIF‐90: Insights from Neural‐Network Molecular Dynamics

Advanced Materials Interfaces, EarlyView.
A DeepMD‐based molecular dynamics framework is developed to investigate water adsorption in ZIF‐90, explicitly accounting for framework flexibility. Results show that atomistic flexibility significantly influences adsorption energetics, diffusion, and structural correlations, yielding nearly constant heat of adsorption and enhanced mobility.
Nick Mackus, Ionut Tranca, Veerapandian Ponnuchamy, Silvia Gaastra‐Nedea   +3 more
wiley   +1 more source

Two‐Way Shape Memory Polymer Composite Gripper for Adaptive Robotic Applications

Advanced Materials Technologies, EarlyView.
A two‐way shape memory polymer (SMP) composite is developed with intrinsic shape‐changing capability driven solely by temperature, eliminating external actuation loads. Embedding the SMP in a low‐stiffness elastomeric matrix enabled reversible transformations during heating and cooling cycles.
Aamna Hameed, Kamran Ahmed Khan
wiley   +1 more source

Memorization in Graph Neural Networks

CoRR
Deep neural networks (DNNs) have been shown to memorize their training data, yet similar analyses for graph neural networks (GNNs) remain largely under-explored. We introduce NCMemo (Node Classification Memorization), the first framework to quantify label memorization in semi-supervised node classification.
Adarsh Jamadandi, Jing Xu, Adam Dziedzic, Franziska Boenisch   +3 more
openaire   +2 more sources

Formal verification of deep neural networks [PDF]

This paper introduces a method for the formal verification of neural networks using a Satisfiability Modulo Theories (SMT) solver. This approach enables the mathematical validation of specific neural network properties, enhancing their predictability. We
Panchuk, B.O.
core