Deep Topological Embedding with Convolutional Neural Networks for Complex Network Classification

Scabini, Leonardo F. S.; Ribas, Lucas Correia; Ribeiro, Eraldo; Bruno, Odemir Martinez

doi:10.1007/978-3-030-97240-0_5

Cited by 3 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Numerous studies found in the literature are devoted to image classification [Costa 2004, Backes, Casanova and Bruno 2009, Backes, Martinez and Bruno 2011, Backes and Bruno 2013, Ribas, Scabini and Bruno 2022, Scabini et al 2017. The volume of work related to vision is due to initial studies that proposed ways to model images as complex networks [Costa 2004].…”

Section: Related Literaturementioning

confidence: 99%

“…A recent work for shape recognition was proposed by SCABINI et al (2017) and after linking the contour points to form a graph, uses the angles formed between the points for the generation of the feature vector. This work was latter applied to classify fish species in [Ribas, Scabini and Bruno 2022] For dynamic textures, where a sequence of textures produces a video, the work of GONÇALVES; MACHADO; BRUNO (2015) is applied. This models the frames and connects them through a graph model to then extract degree measures, considered simple, but achieves good results in the characterization of the images.…”

Section: Related Literaturementioning

confidence: 99%

“…Based on the advantages mentioned above, researchers have successfully used the model for pattern recognition in various applications, such as the classification of static and dynamic textures [Gonçalves, Machado and Bruno 2015], shapes [Ribas, Neiva and Bruno 2019], authorship [Machicao et al 2018], and others. Recently, some works, such as the use of cellular automata in [Miranda, Machicao and Bruno 2016, Zielinski et al 2022, Ribas, Machicao and Bruno 2020, the construction of multidimensional and deep embeddings from networks in [Scabini et al 2022], and the analysis of angles formed in the graph of shapes in [Scabini et al 2017, Ribas, Scabini andBruno 2022], have distinguished themselves from traditional analysis that uses classical statistical metrics such as degree and clustering coefficient to create a one-dimensional graph representation. The recent efforts of some researchers to find novel techniques that overcome the redundant information found in the composition of descriptors based on classical metrics have produced good results in graph classification.…”

Section: Introductionmentioning

confidence: 99%

“…Classifying Complex Networks using Multiscale analysis [Costa et al 2007]. However, alternative methodologies have emerged, including the use of cellular automata, multi-dimensional and deep embeddings, examination of angles formed within CNs, and randomized neural networks for feature extraction [Miranda, Machicao and Bruno 2016, Ribas, Machicao and Bruno 2020, Scabini et al 2017, Zielinski et al 2022, Scabini et al 2022]. These innovations have expanded the scope of complex network analysis.…”

Section: Introductionmentioning

confidence: 99%

“…Furthermore, pattern recognition, combined with complex networks, plays a critical role in identifying entities like shapes , species [Ribas, Scabini and Bruno 2022], and authorship [Marinho, Hirst and Amancio 2016]. This underscores the importance of employing graphs to decipher the structures and patterns of any given organization.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Exploring Complex Networks: Matrix-based and Multiscale Approaches for Pattern Recognition

Neiva

View full text Add to dashboard Cite

Primeiro quero agradecer a minha família que sempre me apoiou em todas as minhas decisões, mesmo sem nem entender o que eu fazia! Um agradecimento especial à meu pai Josué, minha irmã Fabiane, e minha mãe 2 Valderis. Obrigada por cuidarem de mim em dias tão dificeís como foi a minha cirurgia em 2019/2020. Também quero agradecer minha tia Adriana por ser a melhor tia do mundo e um exemplo de mulher forte pra mim! Também sou imensamente grata aos meus avôs Josélio e Marlene por vibrarem e torcerem por mim e terem guardado toda matéria/lista que sai com meu nome no jornal. E como não poderia deixar, agradeço minha mãe por ter esperado tantos dias na escola/kumon/inglês para nos levar sã e salvas pra casa durante alguns anos! Agredeço meu companheiro Etevaldo que durante esses últimos anos de doutorado, enfrentou o dia-a-dia de uma pandemia, com muitas incertezas, medos e ansiedade. Eu nem tenho palavras para agradecer tamanha paciência diária no confinamento! À minhas amigasirmãs, meus apoios durante anos! Tenho certeza que posso contar com vocês sempre e pra sempre, obrigada pelos risos, pelas lágrimas e pela parceria de anos! Agradecimentos especiais são direcionados ao Instituto de Ciências Matemáticas e Computação (ICMC), ao Instituto de Física de São Carlos (IFSC), aos meus parceiros de laboratório do SCG, aos parceiros do projeto RARAS (FMRP-USP) especialmente o professor Dr. Domingos Alves, também ao professor Dr. Antoine Vacavant e, principalmente, ao meu orientador Odemir por toda a motivação e paciência durante esses anos de parceria. Á todos, obrigada.O presente trabalho foi realizado com apoio da Coordenação de Aperfeiçoamento de Pessoal de Nível Superior -Brasil (CAPES) -Código de Financiamento 001."...O crédito pertence ao homem que encontra-se na arena, cuja face está manchada de poeira, suor e sangue; aquele que esforça-se bravamente; que erra, que se depara com um revés após o outro, pois não há esforço sem erros e falhas; aquele que esforça-se para lograr suas ações, que conhece grande entusiasmo, grandes devoções, que se entrega à uma causa nobre; que, no melhor dos casos, conhece no fim o triunfo da realização grandiosa, e quem, que no pior dos casos, se falhar, ao menos falha ousando grandemente, para que seu lugar jamais seja com aquelas frias e tímidas almas que não conhecem vitória ou fracasso." (Citizenship in a Republic -Theodore Roosevelt) RESUMO

show abstract

Section: Related Literaturementioning

confidence: 99%

Section: Related Literaturementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Exploring Complex Networks: Matrix-based and Multiscale Approaches for Pattern Recognition

Neiva

View full text Add to dashboard Cite

show abstract

Patterns and randomness in networks for computer vision: from graphs to neural networks

Scabini

View full text Add to dashboard Cite

To my parents, brothers, grandparents, uncles, and cousins for the familial love and occasional support.To Marilia for her love, companionship, and support, and for all the moments we shared through our journey together.To my advisor, Odemir Martinez Bruno, for the guidance, support, trust, and collaboration for the past seven years. Working together has been an incredible journey and a fundamental part of my professional development.To my co-advisor, Bernard De Baets, firstly for the supervision during my stay at Ghent University, and also for becoming part of the development of this thesis as a whole. It was a great pleasure meeting and working with such an esteemed scientist.Special thanks to my cats, Puck, Juanita, and Gwyn, for their active participation in developing this thesis by constantly walking over the keyboard. They deserve all the credit for any lost characters or weird typos in the text.To all colleagues, professors, and workers from the São Carlos Institute of Physics (IFSC-USP), the Institute of Mathematics and Computer Science (ICMC-USP), the Faculty of Bioscience Engineering (UGent), and the Scientific Computing Group (SCG), who contributed, directly or indirectly, to my academic education and the development of this thesis. I am deeply grateful for the discussions, knowledge exchange, and casual moments that were part of my life during the Ph.D.

show abstract

Dynamical stability and chaos in artificial neural network trajectories along training

Danovski,

Soriano,

Lacasa

2024

Front. Complex Syst.

View full text Add to dashboard Cite

The process of training an artificial neural network involves iteratively adapting its parameters so as to minimize the error of the network’s prediction, when confronted with a learning task. This iterative change can be naturally interpreted as a trajectory in network space–a time series of networks–and thus the training algorithm (e.g., gradient descent optimization of a suitable loss function) can be interpreted as a dynamical system in graph space. In order to illustrate this interpretation, here we study the dynamical properties of this process by analyzing through this lens the network trajectories of a shallow neural network, and its evolution through learning a simple classification task. We systematically consider different ranges of the learning rate and explore both the dynamical and orbital stability of the resulting network trajectories, finding hints of regular and chaotic behavior depending on the learning rate regime. Our findings are put in contrast to common wisdom on convergence properties of neural networks and dynamical systems theory. This work also contributes to the cross-fertilization of ideas between dynamical systems theory, network theory and machine learning.

show abstract

Deep Topological Embedding with Convolutional Neural Networks for Complex Network Classification

Cited by 3 publications

References 21 publications

Exploring Complex Networks: Matrix-based and Multiscale Approaches for Pattern Recognition

Exploring Complex Networks: Matrix-based and Multiscale Approaches for Pattern Recognition

Patterns and randomness in networks for computer vision: from graphs to neural networks

Dynamical stability and chaos in artificial neural network trajectories along training

Contact Info

Product

Resources

About