578 Hits in 6.0 sec

Change Detection in Graph Streams by Learning Graph Embeddings on Constant-Curvature Manifolds [article]

Daniele Grattarola, Daniele Zambon, Cesare Alippi, Lorenzo Livi
2019 arXiv   pre-print
In this paper, we focus on the problem of detecting changes in stationarity in a stream of attributed graphs.  ...  First, via a novel approach based on adversarial learning, we compute graph embeddings by training an autoencoder to represent graphs on CCMs.  ...  Acknowledgements This research is funded by the Swiss National Science Foundation project 200021_172671: "ALPSFORT: A Learning graPh-baSed framework FOr cybeR-physical sysTems".  ... 
arXiv:1805.06299v3 fatcat:v3xpdicj3zg6higeso2gwmij4u

Adversarial Autoencoders with Constant-Curvature Latent Manifolds [article]

Daniele Grattarola, Lorenzo Livi, Cesare Alippi
2019 arXiv   pre-print
Constant-curvature Riemannian manifolds (CCMs) have been shown to be ideal embedding spaces in many application domains, as their non-Euclidean geometry can naturally account for some relevant properties  ...  While a few works in recent literature make use of either hyperspherical or hyperbolic manifolds for different learning tasks, ours is the first unified framework to seamlessly deal with CCMs of different  ...  This research is funded by the Swiss National Science Foundation project ALPSFORT (200021/172671).  ... 
arXiv:1812.04314v2 fatcat:gi53rlqla5gnvavb2cwqslawqe

Error Metrics for Learning Reliable Manifolds from Streaming Data [chapter]

Frank Schoeneman, Suchismit Mahapatra, Varun Chandola, Nils Napp, Jaroslaw Zola
2017 Proceedings of the 2017 SIAM International Conference on Data Mining  
In this paper, we argue that a stable manifold can be learned using only a fraction of the stream, and the remaining stream can be mapped to the manifold in a significantly less costly manner.  ...  We present error metrics that allow us to identify the transition point for a given stream by quantitatively assessing the quality of a manifold learned using Isomap.  ...  Acknowledgment This work was supported by NSF grant CNS:1409551. Access to computing facilities were provided by the UB Center for Computational Research. Usual disclaimers apply.  ... 
doi:10.1137/1.9781611974973.84 dblp:conf/sdm/SchoenemanMCNZ17 fatcat:3yiuyy7yyfdane3htuocje2q5e

Probabilistic modelling of general noisy multi-manifold data sets

M. Canducci, P. Tiño, M. Mastropietro
2021 Artificial Intelligence  
To the best of our knowledge, this work represents the first attempt at detection and modelling of a set of coexisting general noisy manifolds by uniting two aspects of multi-manifold learning: the recovery  ...  We demonstrate the workings of the framework on two synthetic data sets, presenting challenging features for state-of-the-art techniques in Multi-Manifold learning.  ...  PT and MC were supported by the Alan Turing Institute Fellowship 96102.  ... 
doi:10.1016/j.artint.2021.103579 fatcat:hcifg7bgvfcindqria767e76lu

Constant Curvature Graph Convolutional Networks [article]

Gregor Bachmann, Gary Bécigneul, Octavian-Eugen Ganea
2020 arXiv   pre-print
Here, we bridge this gap by proposing mathematically grounded generalizations of graph convolutional networks (GCN) to (products of) constant curvature spaces.  ...  We do this by i) introducing a unified formalism that can interpolate smoothly between all geometries of constant curvature, ii) leveraging gyro-barycentric coordinates that generalize the classic Euclidean  ...  Gary Bécigneul is funded by the Max Planck ETH Center for Learning Systems.  ... 
arXiv:1911.05076v3 fatcat:qfvdahbvxbakpjacnq3jgtcfpq

Unsupervised Detection of Changes in Usage-Phases of a Mobile App

Hoyeol Chae, Ryangkyung Kang, Ho-Sik Seok
2020 Applied Sciences  
This work proposes novel approaches for one of the core functionalities of automated app testing: the detection of changes in usage-phases of a mobile app.  ...  Thus, we propose methods detecting usage-phase changes through object recognition and metrics utilizing graphs and generative models.  ...  In order to represent graphs as points in a non-Euclidean space and detect changes, adversarial training of autoencoders was employed for graph embeddings on constant-curvature manifolds, and change detection  ... 
doi:10.3390/app10103656 fatcat:us5frw3hwncplj7gfhli7pgxze

Online Graph Dictionary Learning [article]

Cédric Vincent-Cuaz, Titouan Vayer, Rémi Flamary, Marco Corneli, Nicolas Courty
2021 arXiv   pre-print
Our approach naturally extends to labeled graphs, and is completed by a novel upper bound that can be used as a fast approximation of Gromov Wasserstein in the embedding space.  ...  Yet, this analysis is not amenable in the context of graph learning, as graphs usually belong to different metric spaces.  ...  This work is supported by the ACA-DEMICS grant of the IDEXLYON, project of the Université de Lyon, PIA operated by ANR-16-IDEX-0005.  ... 
arXiv:2102.06555v2 fatcat:eqg6e7s3lrfe5g5uuihpryjhwq

An Approximate Model for Event Detection from Twitter Data

Aarzoo Dhiman, Durga Toshniwal
2020 IEEE Access  
Earlier research works, addressing these challenges, have tried to capture the contextual information by using the dense vector representations of texts leveraging deep neural word embedding generation  ...  The abundance and real-time availability of Twitter data have proved beneficial in detecting events in various domains such as emergency situations, crime detection, public health, place recommendations  ...  A sphere can be defined as a Riemannian manifold with constant positive curvature.  ... 
doi:10.1109/access.2020.3007004 fatcat:wcszgxqurrbe7lt6osgisqtl2q

High Performance Algorithms for Quantum Gravity and Cosmology [article]

William J. Cunningham
2018 arXiv   pre-print
We then examine the broader applicability of these algorithms to greedy information routing in random geometric graphs embedded in Lorentzian manifolds, which requires us to find new closed-form solutions  ...  The high performance algorithms described herein are the most compact, efficient methods known for representing and analyzing systems modeled well by sets or graphs.  ...  which the number of scalar fields in the low-energy 4D theory, known as Kähler moduli, changes by one.  ... 
arXiv:1805.04463v1 fatcat:xbk732bbzjeann2fpefo5ongdi

Interpretable Signed Link Prediction with Signed Infomax Hyperbolic Graph [article]

Yadan Luo, Zi Huang, Hongxu Chen, Yang Yang, Mahsa Baktashmotlagh
2021 arXiv   pre-print
Most of the prior efforts are devoted to learning node embeddings with graph neural networks (GNNs), which preserve the signed network topology by message-passing along edges to facilitate the downstream  ...  By maximizing the mutual information between edge polarities and node embeddings, one can identify the most representative neighboring nodes that support the inference of edge sign.  ...  ACKNOWLEDGMENTS This work is partially supported by ARC FT130101530, NSFC No. 61628206 and Google PhD Fellowship.  ... 
arXiv:2011.12517v2 fatcat:cbfriruejva2rf3r5of25aitz4

Perspectives on geometric analysis [article]

Shing-Tung Yau
2006 arXiv   pre-print
This is a survey paper on several aspects of differential geometry for the last 30 years, especially in those areas related to non-linear analysis.  ...  Chern who had passed away in December 2004.  ...  bounded from above by a negative constant, are distance decreasing with constants depending only on the bound on the curvature.  ... 
arXiv:math/0602363v2 fatcat:ad4gz5qjhbahjjz2qt6zacogzq

Perspectives on geometric analysis [chapter]

2007 Proceedings of the International Conference on Complex Geometry and Related Fields  
bounded from above by a negative constant, are distance decreasing with constants depending only on the bound on the curvature.  ...  The man whom I admire is on the bank of the river. I go against the stream in quest of him, But the way is difficult and turns to the right. I go down the stream in quest of him, and Lo!  ... 
doi:10.1090/amsip/039/17 fatcat:upv2nxqxw5cbdhtnbixope5unu

A Survey on Knowledge Graphs: Representation, Acquisition and Applications [article]

Shaoxiong Ji and Shirui Pan and Erik Cambria and Pekka Marttinen and Philip S. Yu
2021 IEEE Transactions on Neural Networks and Learning Systems   accepted
In this survey, we provide a comprehensive review of knowledge graph covering overall research topics about 1) knowledge graph representation learning, 2) knowledge acquisition and completion, 3) temporal  ...  In the end, we have a thorough outlook on several promising research directions.  ...  While it fails to capture logical patterns and suffers from constant curvature. Chami et al.  ... 
doi:10.1109/tnnls.2021.3070843 pmid:33900922 arXiv:2002.00388v4 fatcat:4l2yxnf3wbg4zpzdumduvyr4he

Using Hyperbolic Geometry for FG-NET over Distantly Supervised data [article]

Muhammad Asif Ali, Yifang Sun, Bing Li, Wei Wang
2022 arXiv   pre-print
In this paper, we propose to use hyperbolic geometry for FG-NET with the hope that it can help overcoming the noise incurred by distant supervision.  ...  For a manifold with curvature constant K, these operations can be summarized in the equa- 3.5 Complete Model We combine the above-mentioned tion, given below:  ...  Learning surface tity type classification by jointly learning representations and text patterns for a question answering system. In Proceedings label embeddings.  ... 
arXiv:2101.11212v2 fatcat:4tbxz4vnane6fbprrfr3w4lkpq

Representation, Analysis, and Recognition of 3D Humans

Stefano Berretti, Mohamed Daoudi, Pavan Turaga, Anup Basu
2018 ACM Transactions on Multimedia Computing, Communications, and Applications (TOMCCAP)  
Then, a representation is built on lower-level descriptors that model the information embedded in the data.  ...  An increasing importance is also assumed by learning solutions, where hand-crafted descriptors are substituted by deep features that are learned directly from the data.  ...  [40] were among the first to focus on the task of 3D face detection by analyzing the surface curvature.  ... 
doi:10.1145/3182179 fatcat:ds55t4md2na2tibtyg4llerf3q
« Previous Showing results 1 — 15 out of 578 results