Academic literature on the topic 'Node representations'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the lists of relevant articles, books, theses, conference reports, and other scholarly sources on the topic 'Node representations.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Journal articles on the topic "Node representations"
1
Zeng, Liang, Lanqing Li, Ziqi Gao, Peilin Zhao, and Jian Li. "ImGCL: Revisiting Graph Contrastive Learning on Imbalanced Node Classification." Proceedings of the AAAI Conference on Artificial Intelligence 37, no. 9 (June 26, 2023): 11138–46. http://dx.doi.org/10.1609/aaai.v37i9.26319.
Full textAbstract:
Graph contrastive learning (GCL) has attracted a surge of attention due to its superior performance for learning node/graph representations without labels. However, in practice, the underlying class distribution of unlabeled nodes for the given graph is usually imbalanced. This highly imbalanced class distribution inevitably deteriorates the quality of learned node representations in GCL. Indeed, we empirically find that most state-of-the-art GCL methods cannot obtain discriminative representations and exhibit poor performance on imbalanced node classification. Motivated by this observation, we propose a principled GCL framework on Imbalanced node classification (ImGCL), which automatically and adaptively balances the representations learned from GCL without labels. Specifically, we first introduce the online clustering based progressively balanced sampling (PBS) method with theoretical rationale, which balances the training sets based on pseudo-labels obtained from learned representations in GCL. We then develop the node centrality based PBS method to better preserve the intrinsic structure of graphs, by upweighting the important nodes of the given graph. Extensive experiments on multiple imbalanced graph datasets and imbalanced settings demonstrate the effectiveness of our proposed framework, which significantly improves the performance of the recent state-of-the-art GCL methods. Further experimental ablations and analyses show that the ImGCL framework consistently improves the representation quality of nodes in under-represented (tail) classes.
2
Chen, Junhui, Feihu Huang, and Jian Peng. "MSGCN: Multi-Subgraph Based Heterogeneous Graph Convolution Network Embedding." Applied Sciences 11, no. 21 (October 21, 2021): 9832. http://dx.doi.org/10.3390/app11219832.
Full textAbstract:
Heterogeneous graph embedding has become a hot topic in network embedding in recent years and has been widely used in lots of practical scenarios. However, most of the existing heterogeneous graph embedding methods cannot make full use of all the auxiliary information. So we proposed a new method called Multi-Subgraph based Graph Convolution Network (MSGCN), which uses topology information, semantic information, and node feature information to learn node embedding vector. In MSGCN, the graph is firstly decomposed into multiple subgraphs according to the type of edges. Then convolution operation is adopted for each subgraph to obtain the node representations of each subgraph. Finally, the node representations are obtained by aggregating the representation vectors of nodes in each subgraph. Furthermore, we discussed the application of MSGCN with respect to a transductive learning task and inductive learning task, respectively. A node sampling method for inductive learning tasks to obtain representations of new nodes is proposed. This sampling method uses the attention mechanism to find important nodes and then assigns different weights to different nodes during aggregation. We conducted an experiment on three datasets. The experimental results indicate that our MSGCN outperforms the state-of-the-art methods in multi-class node classification tasks.
3
Liu, Dong, Yan Ru, Qinpeng Li, Shibin Wang, and Jianwei Niu. "Semisupervised Community Preserving Network Embedding with Pairwise Constraints." Complexity 2020 (November 10, 2020): 1–14. http://dx.doi.org/10.1155/2020/7953758.
Full textAbstract:
Network embedding aims to learn the low-dimensional representations of nodes in networks. It preserves the structure and internal attributes of the networks while representing nodes as low-dimensional dense real-valued vectors. These vectors are used as inputs of machine learning algorithms for network analysis tasks such as node clustering, classification, link prediction, and network visualization. The network embedding algorithms, which considered the community structure, impose a higher level of constraint on the similarity of nodes, and they make the learned node embedding results more discriminative. However, the existing network representation learning algorithms are mostly unsupervised models; the pairwise constraint information, which represents community membership, is not effectively utilized to obtain node embedding results that are more consistent with prior knowledge. This paper proposes a semisupervised modularized nonnegative matrix factorization model, SMNMF, while preserving the community structure for network embedding; the pairwise constraints (must-link and cannot-link) information are effectively fused with the adjacency matrix and node similarity matrix of the network so that the node representations learned by the model are more interpretable. Experimental results on eight real network datasets show that, comparing with the representative network embedding methods, the node representations learned after incorporating the pairwise constraints can obtain higher accuracy in node clustering task and the results of link prediction, and network visualization tasks indicate that the semisupervised model SMNMF is more discriminative than unsupervised ones.
4
Liu, Li, Weixin Zeng, Zhen Tan, Weidong Xiao, and Xiang Zhao. "Node Importance Estimation with Multiview Contrastive Representation Learning." International Journal of Intelligent Systems 2023 (September 6, 2023): 1–13. http://dx.doi.org/10.1155/2023/5917750.
Full textAbstract:
Node importance estimation is a fundamental task in graph analysis, which can be applied to various downstream applications such as recommendation and resource allocation. However, existing studies merely work under a single view, which neglects the rich information hidden in other aspects of the graph. Hence, in this work, we propose a Multiview Contrastive Representation Learning (MCRL) model to obtain representations of nodes from multiple perspectives and then infer the node importance. Specifically, we are the first to apply the contrastive learning technique to the node importance analysis task, which enhances the expressiveness of graph representations and lays the foundation for importance estimation. Moreover, based on the improved representations, we generate the entity importance score by attentively aggregating the scores from two different views, i.e., node view and node-edge interaction view. We conduct extensive experiments on real-world datasets, and the experimental results show that MCRL outperforms existing methods on all evaluation metrics.
5
Han, Liangzhe, Ruixing Zhang, Leilei Sun, Bowen Du, Yanjie Fu, and Tongyu Zhu. "Generic and Dynamic Graph Representation Learning for Crowd Flow Modeling." Proceedings of the AAAI Conference on Artificial Intelligence 37, no. 4 (June 26, 2023): 4293–301. http://dx.doi.org/10.1609/aaai.v37i4.25548.
Full textAbstract:
Many deep spatio-temporal learning methods have been proposed for crowd flow modeling in recent years. However, most of them focus on designing a spatial and temporal convolution mechanism to aggregate information from nearby nodes and historical observations for a pre-defined prediction task. Different from the existing research, this paper aims to provide a generic and dynamic representation learning method for crowd flow modeling. The main idea of our method is to maintain a continuous-time representation for each node, and update the representations of all nodes continuously according to the streaming observed data. Along this line, a particular encoder-decoder architecture is proposed, where the encoder converts the newly happened transactions into a timestamped message, and then the representations of related nodes are updated according to the generated message. The role of the decoder is to guide the representation learning process by reconstructing the observed transactions based on the most recent node representations. Moreover, a number of virtual nodes are added to discover macro-level spatial patterns and also share the representations among spatially-interacted stations. Experiments have been conducted on two real-world datasets for four popular prediction tasks in crowd flow modeling. The result demonstrates that our method could achieve better prediction performance for all the tasks than baseline methods.
6
Vaghani, Dev. "An Approch for Representation of Node Using Graph Transformer Networks." International Journal for Research in Applied Science and Engineering Technology 11, no. 1 (January 31, 2023): 27–37. http://dx.doi.org/10.22214/ijraset.2023.48485.
Full textAbstract:
Abstract: In representation learning on graphs, graph neural networks (GNNs) have been widely employed and have attained cutting-edge performance in tasks like node categorization and link prediction. However, the majority of GNNs now in use are made to learn node representations on homogenous and fixed graphs. The limits are particularly significant when learning representations on a network that has been incorrectly described or one that is heterogeneous, or made up of different kinds of nodes and edges. This study proposes Graph Transformer Networks (GTNs), which may generate new network structures by finding valuable connections between disconnected nodes in the original graph and learning efficient node representation on the new graphs end-to-end. A basic layer of GTNs called the Graph Transformer layer learns a soft selection of edge types and composite relations to produce meaningful multi-hop connections known as meta-paths. This research demonstrates that GTNs can learn new graph structures from data and tasks without any prior domain expertise and that they can then use convolution on the new graphs to provide effective node representation. GTNs outperformed state-of-the-art approaches that need predefined meta-paths from domain knowledge in all three benchmark node classification tasks without the use of domain-specific graph pre-processing.
7
Li, Haoyang, and Lei Chen. "EARLY: Efficient and Reliable Graph Neural Network for Dynamic Graphs." Proceedings of the ACM on Management of Data 1, no. 2 (June 13, 2023): 1–28. http://dx.doi.org/10.1145/3589308.
Full textAbstract:
Graph neural networks have been widely used to learn node representations for many real-world static graphs. In general, they learn node representations by recursively aggregating information from neighbors. However, graphs in many applications are dynamic, evolving with continuous graph events, such as node feature and graph structure updates. These events require the node representations to be updated accordingly. Currently, due to the real-time requirement, how to efficiently and reliably update node representations under continuous graph events is still an open problem. Recent studies propose two solutions to partially address this problem, but their performance is still limited. First, local-based GNNs only update the nodes directly involved in events, suffering from the quality-deficit issue, since they neglect the other nodes affected by these events. Second, neighbor-sampling GNNs propose to sample neighbors to accelerate neighbor aggregation computations, encountering the neighbor-redundant issue. These sampled neighbors may be similar and cannot reflect the distribution of all neighbors, leading that node representations aggregated on these redundant neighbors may differ from those aggregated on all neighbors. In this paper, we propose an efficient and reliable graph neural network, namely EARLY, to update node representations for dynamic graphs. We first identify the top-k influential nodes that are most affected by graph events. Then, to sample neighbors diversely, we propose a diversity-aware layer-wise sampling technique. We theoretically demonstrate that this technique can decrease the sampling expectation error and learn more reliable node representations. Therefore, the top-k nodes selection and diversity-aware sampling enable EARLY to efficiently update node representations in a reliable way. Extensive experiments on the five real-world graphs demonstrate the effectiveness and efficiency of our proposed EARLY.
8
Liu, Yixin, Yizhen Zheng, Daokun Zhang, Vincent CS Lee, and Shirui Pan. "Beyond Smoothing: Unsupervised Graph Representation Learning with Edge Heterophily Discriminating." Proceedings of the AAAI Conference on Artificial Intelligence 37, no. 4 (June 26, 2023): 4516–24. http://dx.doi.org/10.1609/aaai.v37i4.25573.
Full textAbstract:
Unsupervised graph representation learning (UGRL) has drawn increasing research attention and achieved promising results in several graph analytic tasks. Relying on the homophily assumption, existing UGRL methods tend to smooth the learned node representations along all edges, ignoring the existence of heterophilic edges that connect nodes with distinct attributes. As a result, current methods are hard to generalize to heterophilic graphs where dissimilar nodes are widely connected, and also vulnerable to adversarial attacks. To address this issue, we propose a novel unsupervised Graph Representation learning method with Edge hEterophily discriminaTing (GREET) which learns representations by discriminating and leveraging homophilic edges and heterophilic edges. To distinguish two types of edges, we build an edge discriminator that infers edge homophily/heterophily from feature and structure information. We train the edge discriminator in an unsupervised way through minimizing the crafted pivot-anchored ranking loss, with randomly sampled node pairs acting as pivots. Node representations are learned through contrasting the dual-channel encodings obtained from the discriminated homophilic and heterophilic edges. With an effective interplaying scheme, edge discriminating and representation learning can mutually boost each other during the training phase. We conducted extensive experiments on 14 benchmark datasets and multiple learning scenarios to demonstrate the superiority of GREET.
9
Duan, Wei, Junyu Xuan, Maoying Qiao, and Jie Lu. "Learning from the Dark: Boosting Graph Convolutional Neural Networks with Diverse Negative Samples." Proceedings of the AAAI Conference on Artificial Intelligence 36, no. 6 (June 28, 2022): 6550–58. http://dx.doi.org/10.1609/aaai.v36i6.20608.
Full textAbstract:
Graph Convolutional Neural Networks (GCNs) have been generally accepted to be an effective tool for node representations learning. An interesting way to understand GCNs is to think of them as a message passing mechanism where each node updates its representation by accepting information from its neighbours (also known as positive samples). However, beyond these neighbouring nodes, graphs have a large, dark, all-but forgotten world in which we find the non-neighbouring nodes (negative samples). In this paper, we show that this great dark world holds a substantial amount of information that might be useful for representation learning. Most specifically, it can provide negative information about the node representations. Our overall idea is to select appropriate negative samples for each node and incorporate the negative information contained in these samples into the representation updates. Moreover, we show that the process of selecting the negative samples is not trivial. Our theme therefore begins by describing the criteria for a good negative sample, followed by a determinantal point process algorithm for efficiently obtaining such samples. A GCN, boosted by diverse negative samples, then jointly considers the positive and negative information when passing messages. Experimental evaluations show that this idea not only improves the overall performance of standard representation learning but also significantly alleviates over-smoothing problems.
10
Wang, Renbiao, Fengtai Li, Shuwei Liu, Weihao Li, Shizhan Chen, Bin Feng, and Di Jin. "Adaptive Multi-Channel Deep Graph Neural Networks." Symmetry 16, no. 4 (April 1, 2024): 406. http://dx.doi.org/10.3390/sym16040406.
Full textAbstract:
Graph neural networks (GNNs) have shown significant success in graph representation learning. However, the performance of existing GNNs degrades seriously when their layers deepen due to the over-smoothing issue. The node embedding incline converges to a certain value when GNNs repeat, aggregating the representations of the receptive field. The main reason for over-smoothing is that the receptive field of each node tends to be similar as the layers increase, which leads to different nodes aggregating similar information. To solve this problem, we propose an adaptive multi-channel deep graph neural network (AMD-GNN) to adaptively and symmetrically aggregate information from the deep receptive field. The proposed model ensures that the receptive field of each node in the deep layer is different so that the node representations are distinguishable. The experimental results demonstrate that AMD-GNN achieves state-of-the-art performance on node classification tasks with deep models.
Dissertations / Theses on the topic "Node representations"
1
Marshall, Byron Bennett. "Concept Matching in Informal Node-Link Knowledge Representations." Diss., Tucson, Arizona : University of Arizona, 2005. http://etd.library.arizona.edu/etd/GetFileServlet?file=file:///data1/pdf/etd/azu%5Fetd%5F1145%5F1%5Fm.pdf&type=application/pdf.
Full text
2
Celikkanat, Abdulkadir. "Graph Representation Learning with Random Walk Diffusions." Electronic Thesis or Diss., université Paris-Saclay, 2021. http://www.theses.fr/2021UPASG030.
Full textAbstract:
L'objectif principal de l'Apprentissage de Représentations sur Graphes est de plonger les nœuds dans un espace vectoriel de petite dimension. Dans cette thèse, nous abordons plusieurs enjeux dans le domaine. Tout d'abord, nous étudions comment exploiter l'existence de communautés structurelles locales inhérentes aux graphes tout en apprenant les représentations. Nous apprenons des représentations améliorées de la communauté en combinant les informations latentes avec les représentations. De plus, nous nous concentrons sur l'expressivité des représentations. Nous mettons l'accent sur les distributions de familles exponentielles pour saisir des modèles d'interaction riches. Nous proposons un modèle qui combine les marches aléatoires avec une matrice de factorisation sous forme de noyau. Dans la dernière partie de la thèse, nous étudions des modèles permettant un bon compromis entre efficacité et précision. Nous proposons un modèle évolutif qui calcule des représentations binairesGraph Representation Learning aims to embed nodes in a low-dimensional space. In this thesis, we tackle various challenging problems arising in the field. Firstly, we study how to leverage the inherent local community structure of graphs while learning node representations. We learn enhanced community-aware representations by combining the latent information with the embeddings. Moreover, we concentrate on the expressive- ness of node representations. We emphasize exponential family distributions to capture rich interaction patterns. We propose a model that combines random walks with kernelized matrix factorization. In the last part of the thesis, we study models balancing the trade-off between efficiency and accuracy. We propose a scalable embedding model which computes binary node representations
3
Tibblin, Mark. "A node-based representation of gameplay scripts." Thesis, Uppsala universitet, Institutionen för informationsteknologi, 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-274878.
Full textAbstract:
This project expands on the Unreal Engine 4 by providing an additional tool for it. The tool’s purpose is to convert C# code into the engine’s built in Blueprint system which is a visual scripting kind of language. This thesis will be going through the three big steps necessary to take in order to complete this kind of conversion. It includes parsing and evaluating C# code, creating an Abstract Syntax Tree from it, and generating a Blueprint from the result. Furthermore it will also explore how to structure a graph, as a Blueprint is visualizing through the use of graphs and nodes.
4
Ferreira, Ana Brígida Freitas. "Os nomes da dor: processos de representação nas narrativas de Rui Nunes." Master's thesis, Universidade de Évora, 2009. http://hdl.handle.net/10174/15707.
Full textAbstract:
O fulcro temático da dor é o eixo pendular sobre o qual deverão assentar
motivos que se desenrolam em diversos leitmotiven, e que instituem a problemática da
representação nos corpora narrativos de Rui Nunes seleccionados para esta dissertação
– Cães e A Boca na Cinza. Estas duas obras fundam um meta-paradigma que enforma
um irreversível estranhamento na recepção, e que se pauta por uma labiríntica
compleição diegética e discursiva. Mediante uma narrativa de cariz hiperbólico, importa
equacionar a dor individual, afunilada nas personagens e nas suas intrincadas
deambulações ontológicas, e que se pulveriza em metástases arquetípicas que
desembocam numa acepção colectiva e universal da dor.
A fragmentação do discurso instaura um complexo efeito de palimpsesto com a
fragmentação das categorias da narrativa, desde as personagens até aos lugares onde se
movem ao longo da diegese, num inusitado compromisso meta-reflexivo de causaconsequência
da dor e dos nomes com que se representa.
Ao longo de Cães e de A Boca na Cinza iremos explorar a constante querela
entre as personagens e a memória, ora numa busca incessante por estilhaços
mnemónicos de seres e lugares ausentes (Cães), ora numa contínua recusa de uma
memória que amplia constantemente uma compleição física antípoda do conceito de
humanização (A Boca na Cinza), para apurarmos os distintos confrontos existenciais, e
equacionarmos os processos com que são exaustivamente representados.
A questão da nomeação como representação verbal dos motivos da dor instaura,
igualmente, uma imbricada meta-reflexividade que faz enformar os nomes como
máscaras verosímeis da dor, mas numa permanente dialéctica entre o verbalizável e o
não-verbalizável, toldada pela fragmentação mnemónica a par com o estilhaçamento
referencial.
Julgamos, pois, pertinente, investigar alguns dos processos de representação da
dor na medida em que deixam, no leitor, a veemente chaga da perpetração da dor como
caminho único de humanização, como via única de verbalizar – e, assim, representar
exaustivamente – o humano; ABSTRACT: The pain’s thematic fulcrum is the pendulum on which shall be established the
motives that will develop on several leitmotiven, which will set up the representation
problematic in the narrative corpora of Rui Nunes chosen for this dissertation - The
Cães and the A Boca na Cinza. Both these books found a meta-paradigm, which forms
an irreversible odd feeling in the reception, marked by labyrinthine narrative and
discursive constitution. Taking into account an hyperbolic narrative, it’s important to
question the individual pain, funnelled on the character and its intricate ontological
wanderings, spread in archetypical metastasis that anchor on a collective and universal
meaning of pain.
The speech fragmentation establishes a complex palimpsest effect with the
fragmentation of the narrative categories, from the character to the places where they
move throughout the diegesis, in an unusual meta-reflexive commitment of causeconsequence
of pain, and the names on which it’s represented.
Throughout the Cães and the A Boca na Cinza we will explore the constant
struggle between the characters and the memory, either an endless seek of mnemonic
splinters of absent living beings and places (Cães), or a continuous refusal of a memory
that constantly enlarges an antipodal physical nature of the humanization concept (A
Boca na Cinza), to analyse the distinct existential debates, and to equate the processes
with which they are exhaustingly represented.
The nomination issue as a verbal representation of the pain motives restores, as
well, an imbricate meta-reflexivity that moulds the names as believable masks of pain,
but rolling on a permanent dialectic between the expressable and unexpressable,
overshadowed by the mnemonic fragmentation, hand by hand with the referential
shattering.
We consider relevant to investigate some of the representation processes, for
they leave in the reader a keen wound of pain perpetration as the single trail of
humanization, the only way of expressing – and then to exhaustingly represent – the
human kind.
5
MARANGONI, RICARDO ALEXANDRE. "O NOME E SEUS DESAFIOS: REPRESENTAÇÕES SOCIAIS DO DIRETOR/GESTOR ESCOLAR ACERCA DO SEU AGIR PROFISSIONAL." Universidade Metodista de Sao Paulo, 2017. http://tede.metodista.br/jspui/handle/tede/1660.
Full textAbstract:
Submitted by Noeme Timbo (noeme.timbo@metodista.br) on 2017-06-26T15:40:14Z
No. of bitstreams: 1
Ricardo Alexandre Marangoni2.pdf: 2379464 bytes, checksum: f4794eca1473d65460109527a55a3cb9 (MD5)Made available in DSpace on 2017-06-26T15:40:14Z (GMT). No. of bitstreams: 1 Ricardo Alexandre Marangoni2.pdf: 2379464 bytes, checksum: f4794eca1473d65460109527a55a3cb9 (MD5) Previous issue date: 2017-02-22
This study, attached to the research line “Teacher Formation” of the Post-Graduation in Education of the Metodista University of São Paulo, aimed to investigate the social representations of a group of school managers in a Public School system in the State of São Paulo about their professional act. The vast theoretical production about teaching work testifies the importance of investigating the teachers’ representations, but still they are few studies which approach the school, manager’s act, in a psychol-social perspective of the theory of social representations. In this sense, we based on it and we discussed the relationship between social-practical representations and social-experience representations. The accomplishment of this research covers two moments which interconnect: at the first, the formation of the theoretical corpus, whose approaches express what they should do, that is, the studies which deal with the school manager’s work and the legal guidelines that orientate his/her work. At the second, the data was collected through a questionnaire involving close and open issues. With the first, we knew the participants’ profile and, with the second, we have approached the social representations about their professional action. The gathering expressed what they say they do and what they say what they would like to do. Therefore, we pursued to investigate the following issues. How do they represent themselves? What represents managing a school? What are the social representations about their professional act? How are they built? The results, after the content analysis, we aimed a highly symbolic degree referred to the past, which is manifested in the ground of tensions around a role invested with meaning and sense, which are launched as a challenge. The representations of the researched managers are bound to a politically correct speech and, it was the reason that, at the end of this work, we decided to call upon the mute zone of the representations. With the latter, it became clear to us that the managers do not have a positive social representation to the aspect that pass through their professional act. As well as it has been also possible to confirm the thesis that the contradictions are on the base of the manager’s profession and they guide their work.
Este estudo, vinculado à linha de pesquisa “Formação de Educadores” do Programa de Pós-Graduação em Educação da Universidade Metodista de São Paulo, teve como objetivo investigar as representações sociais de um grupo de gestores escolares da Rede Pública de Ensino do Estado de São Paulo sobre o seu agir profissional. A vasta produção teórica sobre o trabalho docente atesta a importância de se investigar as representações dos professores, mas ainda são poucos os estudos que abordam o agir do gestor, na perspectiva psicossocial da teoria das representações sociais. Nesse sentido, apoiamo-nos nela e discutimos as relações entre representações sociais-práticas e representações sociais-experiências. A realização desta pesquisa abrange dois momentos que se interligam: no primeiro, a formulação do corpus teórico, cujas abordagens expressam o que eles deveriam fazer, ou seja, os estudos que tratam do trabalho do gestor e das diretrizes legais que orientam o seu trabalho. No segundo, os dados foram coletados através de um questionário, envolvendo questões fechadas e abertas. Com a primeira, conhecemos o perfil dos participantes e, com a segunda, nos aproximamos das representações sociais acerca do seu agir profissional. A coleta expressou o que eles dizem fazer e o que eles dizem que gostariam de fazer. Para tanto, procuramos investigar as seguintes questões: Como eles se representam? O que representa dirigir/gerir uma escola? Como desenvolvem o seu trabalho? Quais são as representações sociais acerca do seu agir profissional? Como elas são construídas? Os resultados, após análise de conteúdo, apontaram um grau altamente simbólico atribuído ao passado, que se manifesta no terreno das tensões em torno de um cargo investido de significações e sentidos, que se lançam como desafios. As representações dos gestores pesquisados ligam-se a um discurso politicamente correto e, foi por esse motivo que, ao final deste trabalho, decidimos recorrer a zona muda das representações. Com esta última, ficou-nos evidente que os gestores não possuem uma representação social positiva em relação aos aspectos que perpassam o seu agir profissional. Assim como, também foi possível confirmar a tese de que as contradições estão na base da profissão do gestor e orientam o seu trabalho.
6
Gois, Marcilyanne Moreira. "Representação Nó-profundidade em FPGA para algoritmos evolutivos aplicados ao projeto de redes de larga-escala." Universidade de São Paulo, 2011. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-13012012-102907/.
Full textAbstract:
Diversos problemas do mundo real estão relacionados ao projeto de redes, tais como projeto de circuitos de energia elétrica, roteamento de veículos, planejamento de redes de telecomunicações e reconstrução filogenética. Em geral, esses problemas podem ser modelados por meio de grafos, que manipulam milhares ou milhões de nós (correspondendo às variáveis de entrada), dificultando a obtenção de soluções em tempo real. O Projeto de uma Rede é um problema combinatório, em que se busca encontrar a rede mais adequada segundo um critério como, por exemplo, menor custo, menor caminho e tempo de percurso. A solução desses problemas é, em geral, computacionalmente complexa. Nesse sentido, metaheurísticas como Algoritmos Evolutivos têm sido amplamente investigadas. Diversas pesquisas mostram que o desempenho de Algoritmos Evolutivos para Problemas de Projetos de Redes pode ser aumentado significativamente por meio de representações mais apropriadas. Este trabalho investiga a paralelização da Representação Nó-Profundidade (RNP) em hardware, com o objetivo de encontrar melhores soluções para Problemas de Projetos de Redes. Para implementar a arquitetura de hardware, denominada de HP-RNP (Hardware Parallelized RNP), foi utilizada a tecnologia de FPGA para explorar o alto grau de paralelismo que essa plataforma pode proporcionar. Os resultados experimentais mostraram que o HP-RNP é capaz de gerar e avaliar novas redes em tempo médio limitado por uma constante (O(1))Many problems related to network design can be found in real world applications, such as design of electric circuits, vehicle routing, telecommunication network planning and phylogeny reconstruction. In general, these problems can be modelled using graphs that handle thousands or millions of nodes (input variables), making it hard to obtain solutions in real-time. The Network Design is the combinatorial problem of finding the most suitable network subject to a evaluation criterion as, for example, lower cost, minimal path and time to traverse the network. The solution of those problems is in general computationally complex. Metaheuristics as Evolutionary Algorithms have been widely investigated for such problems. Several researches have shown that the performance of Evolutionary Algorithms for the Network Design Problems can be significantly increased through more appropriated dynamic data structures (encodings). This work investigates the parallelization of Node-Depth Encoding (NDE) in hardware in order to find better solutions for Network Design Problems. To implement the proposed hardware architecture, called HP-NDE (Hardware Parallellized NDE), the FPGA technology was used to explore the high degree of parallelism that such platform can provide. The experimental results have shown that the HP-NDE can generate and evaluate new networks in average time constrained by a constant (O(1))
7
Andersson, Johanna. "Begreppsträning och problemlösning för ett meningsfullt lärande." Thesis, Malmö universitet, Fakulteten för lärande och samhälle (LS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-35537.
Full textAbstract:
Detta självständiga arbete i fördjupningsämnet matematik utgör dels en kunskapsöversikt och dels en empirisk forskningsstudie med utgång i forskningsområdet matematisk begreppsförståelse och matematisk problemlösning och vänder sig i första hand till yrkesverksamma lärare och lärarstuderande. Den producerade kunskapsöversikten är huvudsakligen baserad på forskning som belyser och problematiserar elevers begreppsliga inlärning i samband med matematisk problemlösning. Den centrala frågeställningen som den empiriska forskningsstudien ämnar utreda är Vilken påverkan har begreppsträning på elevers matematiska problemlösningsförmåga? Forskningsstudiens deltagare utgör sju matematikstuderande gymnasieelever och det empiriska materialet (empirin) som dessa deltagare har producerat utgör begreppskartor och lösningsförslag med tillhörande reflektionstext. Två olika analysverktyg som är förankrade i tidigare forskning om begreppsliga representationer och aspekter av två konstruktivistiska vetenskapsteorier tillämpas vid kvalitativ och kvantitativ innehållsanalys av forskningsstudiens empiriska material. Den empiriska forskningsstudiens resultatanalys utgör huvudsakligen en kvantitativ poängbedömning av deltagarnas konstruerade begreppskartor och en kvalitativ bedömning av deltagarnas producerade lösningsförslag med tillhörande reflektionstext. Av den bedrivna forskningsstudiens resultatanalys framgår det att en god förmåga att tillämpa matematikbegrepp vid matematisk problemlösning förutsätter en god förmåga att under begreppsträning beskriva förhållandet mellan motsvarande matematikbegrepp. Vidare tyder studiens resultatanalys på att förmågan att under begreppsträning översätta mellan olika uttrycksformer för att representera matematikbegrepp innebär en god kommunikationsförmåga vid matematisk problemlösning. Avslutningsvis utmynnar diskussionsavsnittet i förslag på hur den empiriska forskningsstudiens resultatanalys via framtida forsknings- och utvecklingsarbeten kan bidra till att nya och effektiva bedömningsverktyg implementeras i matematikundervisningen.This independent project in the major subject mathematics constitutes partly a knowledge review and partly an empirical research study based on the research area of mathematical conceptual understanding and mathematical problem solving and is primarily aimed at teachers and teacher students. The knowledge review produced is mainly based on research that elucidate and problematizes students’ conceptual learning in conjunction with mathematical problem solving. The main question that the empirical research study aims to investigate is What influence does conceptual training in have on students’ mathematical problem solving ability? The participants in the research study constitute seven mathematics students in upper secondary school and the empirical material (empirical evidence) that these participants have produced constitutes concept maps and solution proposals with associated reflection text. Two different analysis tools that are rooted in previous research on conceptual representations and aspects of two constructivist science theories are applied to qualitative and quantitative content analysis of the empirical material of the research study. The empirical research study’s result analysis constitutes mainly a quantitative score assessment of the participants’ constructed concept maps and a qualitative assessment of the participants’ produced solution proposals with associated reflection text. The result analysis of the conducted research study shows that a good ability to apply mathematical concepts in mathematical problem solving requires a good ability to describe the relationship between the corresponding mathematical concepts during concept training. Furthermore, the result analysis of the study indicates that the ability to translate between different forms of expressions to represent mathematical concepts during concept training means a good communication ability in mathematical problem solving. Lastly, the discussion section culminates in suggestions on how the empirical research study’s result analysis through future research and development projects can contribute to the implementation of new and effective assessment tools in mathematical education.
8
Piereti, Saulo Augusto Ribeiro. "Configurador de redes baseado na representação nó-profundidade para efeito de estimação de estado." Universidade de São Paulo, 2007. http://www.teses.usp.br/teses/disponiveis/18/18154/tde-11092007-111648/.
Full textAbstract:
A modelagem em tempo real dos sistemas elétricos de potência (SEP) é extremamente importante para se obter uma operação em tempo real segura e confiável dos mesmos. O configurador de redes (CR) é uma ferramenta fundamental, para modelagem em tempo real dos SEP. A função do CR é determinar, em tempo real, a topologia atual da rede e a correspondente configuração de medidores, no modelo barra-ramo. Para isso, o configurador processa medidas lógicas, que consistem em estados de chaves e disjuntores, bem como dados armazenados em um banco de dados estático, que descreve a conexão dos equipamentos do sistema com as seções de barramento. Em razão de os CRs exigirem algoritmos de busca em um grafo, o desempenho desses algoritmos torna-se fortemente afetado pela forma com que as árvores são computacionalmente representadas. Propõe-se, neste trabalho, um CR tracking, para efeito de estimação de estado, que se baseia em uma nova forma de representar árvores, denominada representação nó-profundidade (RNP). A RNP permite um acesso direto para cada nó de um grafo e pode representar eficientemente árvores (grafos conexos e acíclicos) e florestas (um grafo com uma ou mais árvores). O CR proposto possui as seguintes características: (i) A RNP possibilita uma rápida atualização da topologia da rede, no modelo barra-ramo; (ii) Esta estrutura também permite a realização das etapas configuração de subestação e de rede ao mesmo tempo, diminuindo assim o tempo de processamento necessário para a obtenção do modelo barra-ramo. Para isso, o CR proposto representa cada seção de barramento do SEP como nó de um grafo e usa a RNP e outras duas estruturas de dados, que serão apresentadas no capítulo 5; (iii) Possibilita a associação dos medidores aos componentes do SEP, no modelo barra-ramo, de forma direta. Para isso, o CR proposto usa a RNP e cria barras fictícias para representar os componentes shunt do SEP. Testes realizados comprovam a eficiência e a robustez do configurador proposto tendo em vista os resultados coerentes obtidos para todos os testes, mesmo para os casos em que a mudança nos estados dos dispositivos seccionadores acarretava uma alteração drástica na rede elétrica.On-line models of power system networks have a wide variety of critical uses, covering from security monitoring and control to market operation. Network topology processor (NTP) is a key tool in providing robust and reliable on-line model of power networks. The function of NTP is the determination of the bus/branch topology model (BBTM) of the network and the assignment of metering devices to the components of the BBTM. In order to do this, the NTP processes: logical measurements that consist of switching-device (breakers and switchers) status; as well as a static data-base describing the network connectivity in terms of bus-sections and switching-devices. Since NTPs require search algorithms for graphs, their performance can be drastically affected by the adopted computational graph representation. This work proposes a new tracking NTP for state estimation purposes, that uses a new graph representation named node-depth representation (NDR). This encoding enables a straightforward access to each one of the graph nodes and can efficiently represent trees (acyclic and connected graphs) and forests (a graph with one or more trees). The proposed tracking NTP has the following characteristics: (i) Using NDR for representation of a BBTM of the network, this NTP can track, over time, the changes of the network connectivity in a very direct and fast way; (ii) Processes both steps substation and network configurations in the same time, reducing the CPU time necessary to obtain the BBTM. In order to do this, the proposed NTP represents each bus-section as a graph node and uses NDR and other two data structures, which will be presented in the chapter 5; and (iii) To assign metering devices to the components of the BBTM, in a straightforward way, the proposed NTP creates additional buses, called fictitious buses, to represent shunt devices. The results of several tests have shown the proposed NTP is reliable, fast and suitable for real-time operation.
9
Chesnel, André. "La note bleue : l'expression tsigane dans le jazz à travers la presse anglophone nord-américaine des années 1880 aux années 1940." Thesis, La Rochelle, 2018. http://www.theses.fr/2018LAROF001/document.
Full textAbstract:
Cette recherche s’inscrit dans un ensemble de travaux récents qui ont pour objet la contribution des Européens aux processus de création et d’évolution du jazz. Les migrants européens du XIXe et du début du XXe siècle véhiculent avec eux des singularités culturelles qui marquent la musique américaine. Le rôle des Tsiganes et de leurs représentations outre-Atlantique méritent d’être étudiés. Notre démarche historique s’appuie sur l’établissement d’un corpus issu de la lecture de la presse américaine des années 1880 aux années 1940, confronté à des sources variées dans une perspective pluridisciplinaire (histoire de l’art, musique et géographie). Un plan croisé permet d’étudier dans un premier temps la réception de l’image des Tsiganes dans les divertissements américains, dans la musique romantique et dans le jazz. Nous observons l’omniprésence du thème tsigane, l’apparition d’un véritable mythe et son appropriation américaine. Dans un second temps, nous montrons dans quelle mesure les Tsiganes et leurs musiques contribuent à définir le jazz et ses origines. Enfin, nous analysons dans une troisième partie la circulation des Tsiganes en Amérique et leur installation dans les grandes villes des États-Unis où des musiciens tsiganes jouent du jazzThis academic research work is part of a series of recent studies whose aim is to show to what extent Europeans contributed to the birth of jazz and how they marked the development of this new kind of music. American music bears the stamp of the cultural idiosyncrasies that 19th and 20th century European migrants brought along. It is worthwhile focusing on the role played by Gypsies and the way they were perceived across the Atlantic. The reading of the American press from the 1880s to the 1940s, together with a wide variety of other sources, has provided a solid basis for a historical analysis with a multidisciplinary approach including art history, music and geography. A double-entry framework allows one to first study how Gypsies were perceived and represented in American entertainments, romantic music and jazz. What is noticeable is the omnipresence of the Gypsy theme, the emergence of a myth and its Americanization. Secondly, one can see the way Gypsy musics help define jazz music and give clues as to its origins. The third part is devoted to the Gypsies’ nomadic way of life throughout the United States and their settling in large cities where Gypsy musicians played jazz
10
Souza, Erenice Jesus de. "Escola vista pelas crianças: uma análise das representações sociais da escola na voz (vozes) das crianças." Universidade Presbiteriana Mackenzie, 2011. http://tede.mackenzie.br/jspui/handle/tede/1830.
Full textAbstract:
Made available in DSpace on 2016-03-15T19:42:10Z (GMT). No. of bitstreams: 1
Erenice Jesus de Souza.pdf: 1512808 bytes, checksum: 2c0c7ae1b06e5ec19d17d7965ae72406 (MD5)
Previous issue date: 2011-08-12Fundação de Amparo a Pesquisa do Estado de São Paulo
With the advent of Law No. 11.274/06 a new order is established in the structure and operation of Basic Education in Brazil: the school for nine years duration with the inclusion of children six years old. This new reality leads to reflections on the school as a locus of observation of the relationships established therein and challenges researchers from various fields to devote special attention to understanding the child as a subject of rights and full social actor count for much with the reading their looks and voices that can significantly facilitate the improvement of the quality of education as a subjective public right. In this sense, this research aims to contribute reflections about how children perceive the universe of social representations called school, socio-historical situating it as an object to be investigated. Through the research methodology and qualitative approach in education, we selected a group of children enrolled in the first years of elementary education at a public school in the outskirts of São Paulo and use the techniques of semi-structured interviews with small groups and participant observation obtained major centers of thought and identify the significant elements of content analysis. Based on studies conducted by Pierre Bourdieu, Lev Vygotsky, William Corsaro, Sylvia Lane, Serge Moscovici and Laurence Bardin, among others, this reflection implies a concrete analysis of social representations that the individual has the world around him, the child in relation to school, immersed in a speech quite broad, where the gaps, contradictions and ideology could be detected
Com o advento da lei nº 11.274/06 é instituída uma nova ordem na estrutura e funcionamento da Educação Básica no Brasil: o ensino fundamental de nove anos de duração com a inclusão das crianças de seis anos de idade. Esta nova realidade provoca reflexões acerca da escola enquanto lócus de observação das relações que nela se estabelecem e desafia os pesquisadores das mais diversas áreas a dedicarem especial atenção à compreensão da criança enquanto sujeito de direitos e ator social pleno contando, para tanto, com a leitura de seus olhares e vozes que significativamente podem favorecer a melhoria da qualidade da educação enquanto direito público subjetivo. Neste sentido, a presente pesquisa visa contribuir com as reflexões acerca do modo como as crianças percebem o universo de representações sociais chamado escola, situando-o sociohistoricamente como objeto a ser investigado. Por meio da metodologia de pesquisa de abordagem qualitativa em educação, selecionamos um grupo de crianças matriculadas no primeiro ano do ensino fundamental de uma escola pública da periferia da Grande São Paulo e de uso das técnicas da entrevista semiestruturada com pequenos grupos e da observação participante obtivemos importantes núcleos de pensamento e identificamos elementos significativos à análise de conteúdo. Fundamentada nos estudos realizados por Pierre Bourdieu, Lev Vygotsky, Willian Corsaro, Silvia Lane, Serge Moscovici e Laurence Bardin, entre outros, a presente reflexão implica numa análise concreta das representações sociais que o indivíduo tem do mundo que o rodeia, ou seja, da criança em relação à escola, inserida num discurso bastante amplo, onde as lacunas, as contradições e a ideologia puderam ser detectadas
Books on the topic "Node representations"
1
Grad, Franc. Referendum in sprejemanje nove slovenske ustave. Ljubljana: Inštitut za javno upravo pri pravni fakulteti, 1990.
Find full text
2
Sagey, Elizabeth. The representation of features in non-linear phonology: The articulator node hierarchy. New York: Garland, 1990.
Find full text
3
Film and phenomenology: Toward a realist theory of cinematic representation. Cambridge: Cambridge University Press, 1991.
Find full text
4
Royal Town Planning Institute. Race Relations Panel., ed. Planning authorities and racist representations: A practice guidance note for officers of planning authorities. London: Royal Town Planning Institute, 1996.
Find full text
5
Congress, Trades Union, ed. A speaker's guide to the Fowler review: A TUC briefing note for union representatives. London: Trades Union Congress, 1985.
Find full text
6
Terrorism, National Movement Against, ed. Some tears are less newsworthy: A note on media representation of conflict in Sri Lanka. Colombo: National Movement Against Terrorism, 2006.
Find full text
7
Why we should care about bats: Devastating impact white-nose syndrome is having on one of nature's best pest controllers : oversight hearing before the Subcommittee on Fisheries, Wildlife, Oceans, and Insular Affairs of the Committee on Natural Resources, U.S. House of Representatives, One Hundred Twelfth Congress, first session, Friday, June 24, 2011. Washington: U.S. G.P.O., 2012.
Find full text
8
Arkady, Berenstein, and Retakh Vladimir, eds. Noncommutative birational geometry, representations and combinatorics: AMS Special Session on Noncommutative Birational Geometry, Representations and Cluster Algebras, January 6-7, 2012, Boston, MA. Providence, Rhode Island: American Mathematical Society, 2013.
Find full text
9
1938-, Griffiths Phillip, and Kerr Matthew D. 1975-, eds. Hodge theory, complex geometry, and representation theory. Providence, Rhode Island: Published for the Conference Board of the Mathematical Sciences by the American Mathematical Society, 2013.
Find full text
10
Smith, Steven S. The politics of institutional choice: The formation of the Russian State Duma. Princeton, N.J: Princeton University Press, 2001.
Find full textBook chapters on the topic "Node representations"
1
Aggarwal, Manasvi, and M. N. Murty. "Node Representations." In Machine Learning in Social Networks, 67–88. Singapore: Springer Singapore, 2020. http://dx.doi.org/10.1007/978-981-33-4022-0_4.
Full text
2
Cortés, Xavier, Donatello Conte, and Francesc Serratosa. "Sub-optimal Graph Matching by Node-to-Node Assignment Classification." In Graph-Based Representations in Pattern Recognition, 35–44. Cham: Springer International Publishing, 2019. http://dx.doi.org/10.1007/978-3-030-20081-7_4.
Full text
3
Zheng, Pengkun, Yan Wen, Ming Chen, and Geng Chen. "attr2vec: Learning Node Representations from Attributes of Nodes." In Wireless and Satellite Systems, 546–55. Cham: Springer International Publishing, 2021. http://dx.doi.org/10.1007/978-3-030-69072-4_44.
Full text
4
Dickinson, Peter J., Horst Bunke, Arek Dadej, and Miro Kraetzl. "On Graphs with Unique Node Labels." In Graph Based Representations in Pattern Recognition, 13–23. Berlin, Heidelberg: Springer Berlin Heidelberg, 2003. http://dx.doi.org/10.1007/3-540-45028-9_2.
Full text
5
Jouili, Salim, and Salvatore Tabbone. "Graph Matching Based on Node Signatures." In Graph-Based Representations in Pattern Recognition, 154–63. Berlin, Heidelberg: Springer Berlin Heidelberg, 2009. http://dx.doi.org/10.1007/978-3-642-02124-4_16.
Full text
6
Steenfatt, Niklas, Giannis Nikolentzos, Michalis Vazirgiannis, and Qiang Zhao. "Learning Structural Node Representations on Directed Graphs." In Studies in Computational Intelligence, 132–44. Cham: Springer International Publishing, 2018. http://dx.doi.org/10.1007/978-3-030-05414-4_11.
Full text
7
Fischer, Andreas, Seiichi Uchida, Volkmar Frinken, Kaspar Riesen, and Horst Bunke. "Improving Hausdorff Edit Distance Using Structural Node Context." In Graph-Based Representations in Pattern Recognition, 148–57. Cham: Springer International Publishing, 2015. http://dx.doi.org/10.1007/978-3-319-18224-7_15.
Full text
8
Ye, Rongji, Lixin Cui, Luca Rossi, Yue Wang, Zhuo Xu, Lu Bai, and Edwin R. Hancock. "C2N-ABDP: Cluster-to-Node Attention-Based Differentiable Pooling." In Graph-Based Representations in Pattern Recognition, 70–80. Cham: Springer Nature Switzerland, 2023. http://dx.doi.org/10.1007/978-3-031-42795-4_7.
Full text
9
Escolano, Francisco, Boyan Bonev, and Miguel A. Lozano. "Information-Geometric Graph Indexing from Bags of Partial Node Coverages." In Graph-Based Representations in Pattern Recognition, 52–61. Berlin, Heidelberg: Springer Berlin Heidelberg, 2011. http://dx.doi.org/10.1007/978-3-642-20844-7_6.
Full text
10
Riba, Pau, Josep Lladós, Alicia Fornés, and Anjan Dutta. "Large-Scale Graph Indexing Using Binary Embeddings of Node Contexts." In Graph-Based Representations in Pattern Recognition, 208–17. Cham: Springer International Publishing, 2015. http://dx.doi.org/10.1007/978-3-319-18224-7_21.
Full textConference papers on the topic "Node representations"
1
Wang, Yun, Lun Du, Guojie Song, Xiaojun Ma, Lichen Jin, Wei Lin, and Fei Sun. "Tag2Gauss: Learning Tag Representations via Gaussian Distribution in Tagged Networks." In Twenty-Eighth International Joint Conference on Artificial Intelligence {IJCAI-19}. California: International Joint Conferences on Artificial Intelligence Organization, 2019. http://dx.doi.org/10.24963/ijcai.2019/527.
Full textAbstract:
Keyword-based tags (referred to as tags) are used to represent additional attributes of nodes in addition to what is explicitly stated in their contents, like the hashtags in YouTube. Aside of being auxiliary information for node representation, tags can also be used for retrieval, recommendation, content organization, and event analysis. Therefore, tag representation learning is of great importance. However, to learn satisfactory tag representations is challenging because 1) traditional representation methods generally fail when it comes to representing tags, 2) bidirectional interactions between nodes and tags should be modeled, which are generally not dealt within existing research works. In this paper, we propose a tag representation learning model which takes tag-related node interaction into consideration, named Tag2Gauss. Specifically, since tags represent node communities with intricate overlapping relationships, we propose that Gaussian distributions would be appropriate in modeling tags. Considering the bidirectional interactions between nodes and tags, we propose a tag representation learning model mapping tags to distributions consisting of two embedding tasks, namely Tag-view embedding and Node-view embedding. Extensive evidence demonstrates the effectiveness of representing tag as a distribution, and the advantages of the proposed architecture in many applications, such as the node classification and the network visualization.
2
Chen, Dehua, Tiago Pimentel, Adriano Veloso, and Nivio Ziviani. "Denoising Node Embeddings." In LatinX in AI at Neural Information Processing Systems Conference 2018. Journal of LatinX in AI Research, 2018. http://dx.doi.org/10.52591/lxai201812033.
Full textAbstract:
Node embedding (NE) algorithms capture features of graph’s nodes and represent them in a low dimensional vector space. Graphs are inherently noisy structures, which might reduce the learned representations quality. We propose a novel approach using denoising autoencoders to reduce noise in the learned representation of nodes. Experiments with three state-of-the-art NE algorithms show that our approach effectively reduces noise in a link prediction task.
3
Gao, Hongchang, and Heng Huang. "Deep Attributed Network Embedding." In Twenty-Seventh International Joint Conference on Artificial Intelligence {IJCAI-18}. California: International Joint Conferences on Artificial Intelligence Organization, 2018. http://dx.doi.org/10.24963/ijcai.2018/467.
Full textAbstract:
Network embedding has attracted a surge of attention in recent years. It is to learn the low-dimensional representation for nodes in a network, which benefits downstream tasks such as node classification and link prediction. Most of the existing approaches learn node representations only based on the topological structure, yet nodes are often associated with rich attributes in many real-world applications. Thus, it is important and necessary to learn node representations based on both the topological structure and node attributes. In this paper, we propose a novel deep attributed network embedding approach, which can capture the high non-linearity and preserve various proximities in both topological structure and node attributes. At the same time, a novel strategy is proposed to guarantee the learned node representation can encode the consistent and complementary information from the topological structure and node attributes. Extensive experiments on benchmark datasets have verified the effectiveness of our proposed approach.
4
Liu, Zemin, Yuan Fang, Chenghao Liu, and Steven C. H. Hoi. "Node-wise Localization of Graph Neural Networks." In Thirtieth International Joint Conference on Artificial Intelligence {IJCAI-21}. California: International Joint Conferences on Artificial Intelligence Organization, 2021. http://dx.doi.org/10.24963/ijcai.2021/210.
Full textAbstract:
Graph neural networks (GNNs) emerge as a powerful family of representation learning models on graphs. To derive node representations, they utilize a global model that recursively aggregates information from the neighboring nodes. However, different nodes reside at different parts of the graph in different local contexts, making their distributions vary across the graph. Ideally, how a node receives its neighborhood information should be a function of its local context, to diverge from the global GNN model shared by all nodes. To utilize node locality without overfitting, we propose a node-wise localization of GNNs by accounting for both global and local aspects of the graph. Globally, all nodes on the graph depend on an underlying global GNN to encode the general patterns across the graph; locally, each node is localized into a unique model as a function of the global model and its local context. Finally, we conduct extensive experiments on four benchmark graphs, and consistently obtain promising performance surpassing the state-of-the-art GNNs.
5
Brochier, Robin, Adrien Guille, and Julien Velcin. "Global Vectors for Node Representations." In The World Wide Web Conference. New York, New York, USA: ACM Press, 2019. http://dx.doi.org/10.1145/3308558.3313595.
Full text
6
Zhuo, Wei, and Guang Tan. "Proximity Enhanced Graph Neural Networks with Channel Contrast." In Thirty-First International Joint Conference on Artificial Intelligence {IJCAI-22}. California: International Joint Conferences on Artificial Intelligence Organization, 2022. http://dx.doi.org/10.24963/ijcai.2022/340.
Full textAbstract:
We consider graph representation learning in an unsupervised manner. Graph neural networks use neighborhood aggregation as a core component that results in feature smoothing among nodes in proximity. While successful in various prediction tasks, such a paradigm falls short of capturing nodes' similarities over a long distance, which proves to be important for high-quality learning. To tackle this problem, we strengthen the graph with three types of additional graph views, in which each node is directly linked to a set of nodes with the highest similarity in terms of node features, neighborhood features or local structures. Not restricted by connectivity in the original graph, the generated views provide new and complementary perspectives from which to look at the relationship between nodes. Inspired by the recent success of contrastive learning approaches, we propose a self-supervised method that aims to learn node representations by maximizing the agreement between representations across generated views and the original graph, without the requirement of any label information. We also propose a channel-level contrast approach that greatly reduces computation cost. Extensive experiments on six assortative graphs and three disassortative graphs demonstrate the effectiveness of our approach.
7
Zhu, Hongmin, Fuli Feng, Xiangnan He, Xiang Wang, Yan Li, Kai Zheng, and Yongdong Zhang. "Bilinear Graph Neural Network with Neighbor Interactions." In Twenty-Ninth International Joint Conference on Artificial Intelligence and Seventeenth Pacific Rim International Conference on Artificial Intelligence {IJCAI-PRICAI-20}. California: International Joint Conferences on Artificial Intelligence Organization, 2020. http://dx.doi.org/10.24963/ijcai.2020/202.
Full textAbstract:
Graph Neural Network (GNN) is a powerful model to learn representations and make predictions on graph data. Existing efforts on GNN have largely defined the graph convolution as a weighted sum of the features of the connected nodes to form the representation of the target node. Nevertheless, the operation of weighted sum assumes the neighbor nodes are independent of each other, and ignores the possible interactions between them. When such interactions exist, such as the co-occurrence of two neighbor nodes is a strong signal of the target node's characteristics, existing GNN models may fail to capture the signal. In this work, we argue the importance of modeling the interactions between neighbor nodes in GNN. We propose a new graph convolution operator, which augments the weighted sum with pairwise interactions of the representations of neighbor nodes. We term this framework as Bilinear Graph Neural Network (BGNN), which improves GNN representation ability with bilinear interactions between neighbor nodes. In particular, we specify two BGNN models named BGCN and BGAT, based on the well-known GCN and GAT, respectively. Empirical results on three public benchmarks of semi-supervised node classification verify the effectiveness of BGNN --- BGCN (BGAT) outperforms GCN (GAT) by 1.6% (1.5%) in classification accuracy. Codes are available at: https://github.com/zhuhm1996/bgnn.
8
Zhang, Daokun, Jie Yin, Xingquan Zhu, and Chengqi Zhang. "User Profile Preserving Social Network Embedding." In Twenty-Sixth International Joint Conference on Artificial Intelligence. California: International Joint Conferences on Artificial Intelligence Organization, 2017. http://dx.doi.org/10.24963/ijcai.2017/472.
Full textAbstract:
This paper addresses social network embedding, which aims to embed social network nodes, including user profile information, into a latent low-dimensional space. Most of the existing works on network embedding only consider network structure, but ignore user-generated content that could be potentially helpful in learning a better joint network representation. Different from rich node content in citation networks, user profile information in social networks is useful but noisy, sparse, and incomplete. To properly utilize this information, we propose a new algorithm called User Profile Preserving Social Network Embedding (UPP-SNE), which incorporates user profile with network structure to jointly learn a vector representation of a social network. The theme of UPP-SNE is to embed user profile information via a nonlinear mapping into a consistent subspace, where network structure is seamlessly encoded to jointly learn informative node representations. Extensive experiments on four real-world social networks show that compared to state-of-the-art baselines, our method learns better social network representations and achieves substantial performance gains in node classification and clustering tasks.
9
Gao, Li, Hong Yang, Chuan Zhou, Jia Wu, Shirui Pan, and Yue Hu. "Active Discriminative Network Representation Learning." In Twenty-Seventh International Joint Conference on Artificial Intelligence {IJCAI-18}. California: International Joint Conferences on Artificial Intelligence Organization, 2018. http://dx.doi.org/10.24963/ijcai.2018/296.
Full textAbstract:
Most of current network representation models are learned in unsupervised fashions, which usually lack the capability of discrimination when applied to network analysis tasks, such as node classification. It is worth noting that label information is valuable for learning the discriminative network representations. However, labels of all training nodes are always difficult or expensive to obtain and manually labeling all nodes for training is inapplicable. Different sets of labeled nodes for model learning lead to different network representation results. In this paper, we propose a novel method, termed as ANRMAB, to learn the active discriminative network representations with a multi-armed bandit mechanism in active learning setting. Specifically, based on the networking data and the learned network representations, we design three active learning query strategies. By deriving an effective reward scheme that is closely related to the estimated performance measure of interest, ANRMAB uses a multi-armed bandit mechanism for adaptive decision making to select the most informative nodes for labeling. The updated labeled nodes are then used for further discriminative network representation learning. Experiments are conducted on three public data sets to verify the effectiveness of ANRMAB.
10
Wang, Qixiang, Shanfeng Wang, Maoguo Gong, and Yue Wu. "Feature Hashing for Network Representation Learning." In Twenty-Seventh International Joint Conference on Artificial Intelligence {IJCAI-18}. California: International Joint Conferences on Artificial Intelligence Organization, 2018. http://dx.doi.org/10.24963/ijcai.2018/390.
Full textAbstract:
The goal of network representation learning is to embed nodes so as to encode the proximity structures of a graph into a continuous low-dimensional feature space. In this paper, we propose a novel algorithm called node2hash based on feature hashing for generating node embeddings. This approach follows the encoder-decoder framework. There are two main mapping functions in this framework. The first is an encoder to map each node into high-dimensional vectors. The second is a decoder to hash these vectors into a lower dimensional feature space. More specifically, we firstly derive a proximity measurement called expected distance as target which combines position distribution and co-occurrence statistics of nodes over random walks so as to build a proximity matrix, then introduce a set of T different hash functions into feature hashing to generate uniformly distributed vector representations of nodes from the proximity matrix. Compared with the existing state-of-the-art network representation learning approaches, node2hash shows a competitive performance on multi-class node classification and link prediction tasks on three real-world networks from various domains.
Reports on the topic "Node representations"
1
Galiani, Sebastian, and Gustavo Torrens. Why Not Taxation and Representation? A Note on the American Revolution. Cambridge, MA: National Bureau of Economic Research, October 2016. http://dx.doi.org/10.3386/w22724.
Full text
2
Gohin, Alex, and Thomas Hertel. A Note on the CES Functional Form and Its Use in the GTAP Model. GTAP Research Memoranda, October 2003. http://dx.doi.org/10.21642/gtap.rm02.
Full textAbstract:
The purpose of this note is to provide an exhaustive reference for those interested in learning more about the Constant Elasticity of Substitution (CES) function and its use in the representation of producer behavior in the GTAP model. Particular attention is paid to the role of technical change variables and their effect on cost minimizing demands and input shares. This note is divided into three sections. In the first section, the basic cost minimization problem is laid out and conditional factor demands, as well as the unit cost function, are derived. In section two, this system of equations is expressed in terms of proportional changes, as currently specified in GTAP. This greatly facilitates decomposition of predicted changes in demands and costs between three effects, namely expansion, substitution, and technical change effects. Section two also shows the relationship between changes in cost shares and changes in prices and factor-biased technical change variables. Finally, section three relates these derivations to the notation employed in GTAP.
3
Astesiano, Gastón, Carolina Lembo, Ancor Suárez-Alemán, Cristina Simón Morientes, Paula Castillo Martínez, Ana Beatriz Araújo, and Jaime Hurtado. Concept Note for the PPP Talk panels on Driving Inclusion and Measuring Impact. Inter-American Development Bank, May 2023. http://dx.doi.org/10.18235/0004863.
Full textAbstract:
The Inter-American Development Bank (IDB) holds the PPP Americas every two years in partnership with a national or subnational government. The regional forum brings together top professionals and public and private-sector representatives from Latin America and the Caribbean (LAC) to discuss groundbreaking topics and exchange ideas on planning, structuring, and managing public-private partnerships (PPPs). For the 2023 conference, the IDB is holding three preparatory events - the PPP Talks before the main event. During the PPP Talks, invited experts will lay the groundwork for the discussions to be held during the conference. The third PPP Talk will consist of two panels on May 4th, 2023: Driving Inclusion and Measuring Impact, both topics are part of the thematic agenda of PPP Americas 2023. This Concept Note provides the conceptual framework of the two themes discussed during these panels. In addition, this note offers descriptions of key concepts, as well as the main opportunities and challenges that countries in Latin America and the Caribbean face in each area.
4
Astesiano, Gastón, Carolina Lembo, Cristina Simón Morientes, and Paula Castillo Martínez. Concept Note for the PPP Talk panels on Climate Investment and Digital Transformation. Inter-American Development Bank, November 2022. http://dx.doi.org/10.18235/0004573.
Full textAbstract:
The Inter-American Development Bank (IDB) holds the PPP Americas every two years in partnership with a national or subnational government. The regional forum brings together top professionals and public and private-sector representatives from Latin America and the Caribbean (LAC) to discuss groundbreaking topics and exchange ideas on planning, structuring, and managing public-private partnerships (PPPs). For the 2023 conference, the IDB is holding three preparatory events - the PPP Talks before the main event. During the PPP Talks, invited experts will lay the groundwork for the discussions to be held during the conference. The first PPP Talk will consist of two panels on December 1st, 2022: Climate Investment and Digital Transformation both topics are part of the thematic agenda of PPP Americas 2023. This Concept Note provides the conceptual framework of the two themes discussed during these panels. In addition, this note offers descriptions of key concepts, as well as the main opportunities and challenges that countries in Latin America and the Caribbean face in each area.
5
Gidron, Yotam, Lydia Ayikoru, Geoffrey Owino, Katja Kjar-Levin, and Marie Sophie Pettersson. Oxfam's Engagement with Refugee-led Organisations in West Nile (Uganda): Lessons on opportunities and challenges. Oxfam International, December 2023. http://dx.doi.org/10.21201/2023.000006.
Full textAbstract:
Since 2018, Oxfam in Uganda has been supporting South Sudanese refugee-led organizations (RLOs) with funding for a range of peacebuilding, COVID-19 and livelihood activities. Drawing on this experience, this briefing note summarizes some of the most important lessons learned in the working relationship of Oxfam and Uganda-based RLOs. Practical recommendations are also drawn that may be useful for international actors working with RLOs in various contexts, focusing on how to nurture meaningful partnerships, including with regards to funding, capacity building, representation of women and youth, advocacy, knowledge sharing and visibility.
6
Jablonski, David. DTRT57-09-C-10046 Digital Imaging of Pipeline Mechanical Damage and Residual Stress. Chantilly, Virginia: Pipeline Research Council International, Inc. (PRCI), February 2010. http://dx.doi.org/10.55274/r0011872.
Full textAbstract:
The purpose of this program was to enhance the characterization of mechanical damage in pipelines through the application of digital eddy current imaging. Lift-off maps can be used to develop quantitative representations of mechanical damage and magnetic permeability maps can be used to determine residual stress patterns around mechanical damage sites. Note that magnetic permeability is also affected by microstructure variations due to plastic deformation and plowing. High-resolution digital images provide an opportunity for automated analysis of both size and shape of damage and a permanent archival record that can be compared against future measurements to detect changes in size or shape of the damage. Also, multiple frequency measurements will enable volumetric and even through-wall imaging at mechanical damage sites to support further risk assessment efforts.
7
Rojas, Eduardo. Monitoring Urban Markets: The Experience of Latin American Cities. Inter-American Development Bank, May 1997. http://dx.doi.org/10.18235/0008951.
Full textAbstract:
This Urban Sector present Dissemination Note reports on the presentations made and discussions held in the Seminar organized by the Lincoln Institute of Land Policy of Cambridge, MA, on the Latin American experience in monitoring urban land markets. The Seminar gathered 10 representatives of private firms that, as part of their business, systematically gather data and sale information about urban real estate markets. Systematic information about urban land and real estate transactions play a significant role in improving the operation of urban markets. Further, this information is a valuable input for public decision making concerning infrastructure investment, zoning and other land use planning decisions, and for scientific research on urban issues.
8
Gurgel, Angelo, Gilbert Metcalf, and John Reilly. Comparing Tax Rates Using OECD and GTAP6 Data. GTAP Research Memoranda, May 2006. http://dx.doi.org/10.21642/gtap.rm07.
Full textAbstract:
In prior research, Babiker, Metcalf, and Reilly (2003) computed capital and labor tax rates to augment GTAP 4 data that did not comprehensively cover taxes. Recently, we updated these tax rates to 2001 to match the GTAP 6 base year. We realize there was an attempt to improve representation of taxes in GTAP 6, but as admitted by GTAP this was as yet incomplete. For those who might require a more complete tax accounting we have followed an approach that has the benefit of, at least in the aggregate, reconciliation with widely available tax receipt data. This note compares the tax rates derived for the OECD countries using this approach to those provided by the method GTAP uses. Those who are interested in economic impacts of fiscal measures--such as assuring revenue neutrality by raising tax rates in the face of a policy that reduces economic activity--may want to consider amending the tax data in GTAP. Alternatively, it may be useful to discuss the differences we note to as part of on-going efforts to improve data reported in GTAP.
9
Zanoni, Wladimir, Jimena Romero, Nicolás Chuquimarca, and Emmanuel Abuelafia. Dealing with Hard-to-Reach Populations in Panel Data: Respondent-Driven Survey (RDS) and Attrition. Inter-American Development Bank, October 2023. http://dx.doi.org/10.18235/0005194.
Full textAbstract:
Hidden populations, such as irregular migrants, often elude traditional probabilistic sampling methods. In situations like these, chain-referral sampling techniques like Respondent-Driven Surveys (RDS) offer an effective solution. RDS, a variant of network sampling sometimes referred to as “snowball” sampling, estimates weights based on the network structures of friends and acquaintances formed during the sampling process. This ensures the samples are representative of the larger population. However, one significant limitation of these methods is the rigidity of the weights. When faced with participant attrition, recalibrating these weights to ensure continued representation poses a challenge. This technical note introduces a straightforward methodology to account for such attrition. Its applicability is demonstrated through a survey targeting Venezuelan migrants in Ecuador and Peru.
10
Vas, Dragos, Elizabeth Corriveau, Lindsay Gaimaro, and Robyn Barbato. Challenges and limitations of using autonomous instrumentation for measuring in situ soil respiration in a subarctic boreal forest in Alaska, USA. Engineer Research and Development Center (U.S.), December 2023. http://dx.doi.org/10.21079/11681/48018.
Full textAbstract:
Subarctic and Arctic environments are sensitive to warming temperatures due to climate change. As soils warm, soil microorganisms break down carbon and release greenhouse gases such as methane (CH₄) and carbon dioxide (CO₂). Recent studies examining CO₂ efflux note heterogeneity of microbial activity across the landscape. To better understand carbon dynamics, our team developed a predictive model, Dynamic Representation of Terrestrial Soil Predictions of Organisms’ Response to the Environment (DRTSPORE), to estimate CO₂ efflux based on soil temperature and moisture estimates. The goal of this work was to acquire respiration rates from a boreal forest located near the town of Fairbanks, Alaska, and to provide in situ measurements for the future validation effort of the DRTSPORE model estimates of CO₂ efflux in cold climates. Results show that soil temperature and seasonal soil thaw depth had the greatest impact on soil respiration. However, the instrumentation deployed significantly altered the soil temperature, moisture, and seasonal thaw depth at the survey site and very likely the soil respiration rates. These findings are important to better understand the challenges and limitations associated with the in situ data collection used for carbon efflux modeling and for estimating soil microbial activity in cold environments.