4 research outputs found
Corrigendum to “Exploring the potential of BBNCo glasses: Physical, optical, and radiation shielding analysis” [Opt. Mater. 142 (2023) 113976] (Optical Materials (2023) 142, (S0925346723005487), (10.1016/j.optmat.2023.113976))
No full textThe authors regret M.A.M. Uosifa, Shams A.M. Issab, c, A.S. Abouhaswad,e, A.M.A. Mostafaa, Ali Attaa, Hesham M.H. Zakalyc,f,g,*** > aPhysics Department, College of Science, Jouf University, P.O. Box: 2014, Sakaka, Saudi Arabia bDepartment of Physics, Faculty of Science, University of Tabuk, Tabuk, Saudi Arabia cPhysics Department, Faculty of Science, Al-Azhar University, Assiut, 71452, Egypt dPhysics Department, Faculty of Science, Menoufia University, Shebin El-Koom, Menoufia, Egypt eInstitute of Natural Science and Mathematics, Ural Federal University, Ekaterinburg 620002, Russian Federation fIstinye University, Faculty of Engineering and Natural Sciences, Computer Engineering Department, Istanbul, 34396, Turkey gInstitute of Physics and Technology, Ural Federal University, 19 Mira St., 620002, Yekaterinburg, Russia *Corresponding author. Physics Department, College of Science, Jouf University, P.O. Box: 2014, Sakaka, Saudi Arabia. **Corresponding author. Department of Physics, Faculty of Science, University of Tabuk, Tabuk, Saudi Arabia. *** Corresponding author. Physics Department, Faculty of Science, Al-Azhar University, Assiut, 71452, Egypt. E-mail addresses: [email protected] (M.A.M. Uosif), [email protected] (S.A.M. Issa), [email protected] (H.M.H. Zakaly) The authors would like to apologise for any inconvenience caused. © 2023 Elsevier B.V
GraphMix: Improved Training of GNNs for Semi-Supervised Learning
No full textWe present GraphMix, a regularization method for Graph Neural Network based semi-supervised object classification, whereby we propose to train a fully-connected network jointly with the graph neural network via parameter sharing and interpolation-based regularization. Further, we provide a theoretical analysis of how GraphMix improves the generalization bounds of the underlying graph neural network, without making any assumptions about the "aggregation" layer or the depth of the graph neural networks. We experimentally validate this analysis by applying GraphMix to various architectures such as Graph Convolutional Networks, Graph Attention Networks and Graph-U-Net. Despite its simplicity, we demonstrate that GraphMix can consistently improve or closely match state-of-the-art performance using even simpler architectures such as Graph Convolutional Networks, across three established graph benchmarks: Cora, Citeseer and Pubmed citation network datasets, as well as three newly proposed datasets: Cora-Full, Co-author-CS and Co-author-Physics
Adaptive Multi-layer Contrastive Graph Neural Networks
No full textWe present Adaptive Multi-layer Contrastive Graph Neural Networks (AMC-GNN),
a self-supervised learning framework for Graph Neural Network, which learns
feature representations of sample data without data labels. AMC-GNN generates
two graph views by data augmentation and compares different layers' output
embeddings of Graph Neural Network encoders to obtain feature representations,
which could be used for downstream tasks. AMC-GNN could learn the importance
weights of embeddings in different layers adaptively through the attention
mechanism, and an auxiliary encoder is introduced to train graph contrastive
encoders better. The accuracy is improved by maximizing the representation's
consistency of positive pairs in the early layers and the final embedding
space. Our experiments show that the results can be consistently improved by
using the AMC-GNN framework, across four established graph benchmarks: Cora,
Citeseer, Pubmed, DBLP citation network datasets, as well as four newly
proposed datasets: Co-author-CS, Co-author-Physics, Amazon-Computers,
Amazon-Photo.Comment: 16 pages,7 figure
An introduction to tensors and group theory for physicists
No full textThe second edition of this highly praised textbook provides an introduction to tensors, group theory, and their applications in classical and quantum physics. Both intuitive and rigorous, it aims to demystify tensors by giving the slightly more abstract but conceptually much clearer definition found in the math literature, and then connects this formulation to the component formalism of physics calculations. New pedagogical features, such as new illustrations, tables, and boxed sections, as well as additional “invitation” sections that provide accessible introductions to new material, offer increased visual engagement, clarity, and motivation for students. Part I begins with linear algebraic foundations, follows with the modern component-free definition of tensors, and concludes with applications to physics through the use of tensor products. Part II introduces group theory, including abstract groups and Lie groups and their associated Lie algebras, then intertwines this material with that of Part I by introducing representation theory. Examples and exercises are provided in each chapter for good practice in applying the presented material and techniques. Prerequisites for this text include the standard lower-division mathematics and physics courses, though extensive references are provided for the motivated student who has not yet had these. Advanced undergraduate and beginning graduate students in physics and applied mathematics will find this textbook to be a clear, concise, and engaging introduction to tensors and groups. Reviews of the First Edition “[P]hysicist Nadir Jeevanjee has produced a masterly book that will help other physicists understand those subjects [tensors and groups] as mathematicians understand them… From the first pages, Jeevanjee shows amazing skill in finding fresh, compelling words to bring forward the insight that animates the modern mathematical view…[W]ith compelling force and clarity, he provides many carefully worked-out examples and well-chosen specific problems… Jeevanjee’s clear and forceful writing presents familiar cases with a freshness that will draw in and reassure even a fearful student. [This] is a masterpiece of exposition and explanation that would win credit for even a seasoned author.” —Physics Today "Jeevanjee’s [text] is a valuable piece of work on several counts, including its express pedagogical service rendered to fledgling physicists and the fact that it does indeed give pure mathematicians a way to come to terms with what physicists are saying with the same words we use, but with an ostensibly different meaning. The book is very easy to read, very user-friendly, full of examples...and exercises, and will do the job the author wants it to do with style.” —MAA Review
