Filters








32,187 Hits in 4.8 sec

Hyper-Path-Based Representation Learning for Hyper-Networks

Jie Huang, Xin Liu, Yangqiu Song
2019 Proceedings of the 28th ACM International Conference on Information and Knowledge Management - CIKM '19  
Network representation learning has aroused widespread interests in recent years.  ...  Then we propose a new concept called hyper-path and design hyper-path-based random walks to preserve the structural information of hyper-networks according to the analysis of the indecomposability.  ...  -Metapath2vec [9] Metapath2vec is a meta-path-based representations learning model for heterogeneous information networks.  ... 
doi:10.1145/3357384.3357871 dblp:conf/cikm/00090S19 fatcat:ozle2gfc3nhpxedil3oednbo5a

EPB892599 Supplemental Material - Supplemental material for Graph input representations for machine learning applications in urban network analysis

Alessio Pagani, Abhinav Mehrotra, Mirco Musolesi
2020 Figshare  
Supplemental material, EPB892599 Supplemental Material for Graph input representations for machine learning applications in urban network analysis by Alessio Pagani, Abhinav Mehrotra and Mirco Musolesi  ...  More specifically, the models based on Deep Neural Networks are more robust to noise.  ...  This shows that the proposed input representations are suitable as input for training models using different ML algorithms and that they effective represent characteristics of the original network paths  ... 
doi:10.25384/sage.11948520.v1 fatcat:4zxhh7mthfcr5bwxypkv7x7lxa

Memory-Efficient Learned Image Compression with Pruned Hyperprior Module [article]

Ao Luo, Heming Sun, Jinming Liu, Jiro Katto
2022 arXiv   pre-print
Learned Image Compression (LIC) gradually became more and more famous in these years. The hyperprior-module-based LIC models have achieved remarkable rate-distortion performance.  ...  In our research, we found the hyperprior module is not only highly over-parameterized, but also its latent representation contains redundant information.  ...  In recent several years, deep-learning based image compression (Learned Image Compression, LIC) methods take use of neural network, which has non-linear activation, to compress the images.  ... 
arXiv:2206.10083v1 fatcat:th2iqver55hxbf5x72232e7dpe

LSTM Iteration Networks: An Exploration of Differentiable Path Finding

Lisa Lee, Emilio Parisotto, Devendra Singh Chaplot, Ruslan Salakhutdinov
2018 International Conference on Learning Representations  
We do this by designing a new path planning architecture called the LSTM-Iteration Network, which achieves better performance than VINs in metrics such as success rate, training stability, and sensitivity  ...  Our motivation is to scale value iteration to larger environments without a huge increase in computational demand, and fix the problems inherent to Value Iteration Networks (VIN) such as spatial invariance  ...  These results were attained using iteration count K = 20 for all models, filter size F = 3 for VIN and Hyper-VIN, and F = 11 for LSTMIN.  ... 
dblp:conf/iclr/LeePCS18 fatcat:5u5um5fcgbhghlazw5ecqyc4kq

Graph representation learning: a survey

Fenxiao Chen, Yun-Cheng Wang, Bin Wang, C.-C. Jay Kuo
2020 APSIPA Transactions on Signal and Information Processing  
Research on graph representation learning has received great attention in recent years since most data in real-world applications come in the form of graphs.  ...  Various graph embedding techniques have been developed to convert the raw graph data into a low-dimensional vector representation while preserving the intrinsic graph properties.  ...  Hyper-graph representation learning provides a good tool for social network modeling, and it has been a hot research topic nowadays.  ... 
doi:10.1017/atsip.2020.13 fatcat:lirq3kp25jfilgkf66u2rlkhky

Knowledge-Enhanced Top-K Recommendation in Poincaré Ball [article]

Chen Ma, Liheng Ma, Yingxue Zhang, Haolun Wu, Xue Liu, Mark Coates
2021 arXiv   pre-print
Furthermore, a hyperbolic attention network is employed to determine the relative importances of neighboring entities of a certain item.  ...  In addition, we propose an adaptive and fine-grained regularization mechanism to adaptively regularize items and their neighboring representations.  ...  Due to their ability to learn salient representations, (deep) neural network-based methods (He et al. 2017; Sun et al. 2019; are also adopted.  ... 
arXiv:2101.04852v2 fatcat:lipdisr26jcqna3nfipdytfbw4

Identification of Melanoma from Hyperspectral Pathology Image using 3D Convolutional Networks

Qian Wang, Li Sun, Yan Wang, Mei Zhou, Menghan Hu, Jiangang Chen, Ying Wen, Qingli Li
2020 IEEE Transactions on Medical Imaging  
These findings demonstrated the ability of the Hyper-net for assisting pathologists in diagnosis of melanoma based on hyperspectral pathology images.  ...  To make the best use of spectral properties of three dimensional hyperspectral data, we proposed a 3D fully convolutional network named Hyper-net to segment melanoma from hyperspectral pathology images  ...  encoder network by reconstructing nonlinear global-to-local representation.  ... 
doi:10.1109/tmi.2020.3024923 pmid:32956043 fatcat:3gcica5m5fdkndq53c43b3rquy

Few-Shot Semantic Relation Prediction across Heterogeneous Graphs [article]

Pengfei Ding, Yan Wang, Guanfeng Liu, Xiaofang Zhou
2022 arXiv   pre-print
Targeting this novel and challenging problem, in this paper, we propose a Meta-learning based Graph neural network for Semantic relation prediction, named MetaGS.  ...  Thirdly, using the well-initialized two-view graph neural network and hyper-prototypical network, MetaGS can effectively learn new semantic relations from different graphs while overcoming the limitation  ...  network to learn representations of semantic relations by aggregating information of these subgraphs.  ... 
arXiv:2207.05068v1 fatcat:awbc6tiyefanbnp7lo7bozdaom

Identification of Melanoma From Hyperspectral Pathology Image Using 3D Convolutional Networks

Qian Wang, Li Sun, Yan Wang, Mei Zhou, Menghan Hu, Jiangang Chen, Ying Wen, Qingli Li
2021 IEEE Transactions on Medical Imaging  
These findings demonstrated the ability of the Hyper-net for assisting pathologists in diagnosis of melanoma based on hyperspectral pathology images.  ...  To make the best use of spectral properties of three dimensional hyperspectral data, we proposed a 3D fully convolutional network named Hyper-net to segment melanoma from hyperspectral pathology images  ...  On the other hand, deep learning based methods have been applied on hyperspectral imaging especially on remote sensing datasets where the biggest concern is the limited data for training 3D CNN networks  ... 
doi:10.1109/tmi.2020.3024923 fatcat:uxuiv5zisrdivn6s36gfgkjem4

Hyper-Pairing Network for Multi-Phase Pancreatic Ductal Adenocarcinoma Segmentation [article]

Yuyin Zhou, Yingwei Li, Zhishuai Zhang, Yan Wang, Angtian Wang, Elliot Fishman, Alan Yuille, Seyoun Park
2019 arXiv   pre-print
The proposed approach consists of a dual path network where the two parallel streams are interconnected with hyper-connections for intensive information exchange.  ...  To this end, we present Hyper-Pairing Network (HPN), a 3D fully convolution neural network which effectively integrates information from different phases.  ...  This work was supported by the Lustgarten Foundation for Pancreatic Cancer Research. We thank Fengze Liu, Yingda Xia, Qihang Yu and Zhuotun Zhu for instructive discussions.  ... 
arXiv:1909.00906v1 fatcat:7jwcr7b3vvdfvfrdrdzwqiemcy

Knowledge-Enhanced Top-K Recommendation in Poincaré Ball

Chen Ma, Liheng Ma, Yingxue Zhang, Haolun Wu, Xue Liu, Mark Coates
2021 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Furthermore, a hyperbolic attention network is employed to determine the relative importances of neighboring entities of a certain item.  ...  In addition, we propose an adaptive and fine-grained regularization mechanism to adaptively regularize items and their neighboring representations.  ...  Due to their ability to learn salient representations, (deep) neural network-based methods (He et al. 2017; Sun et al. 2019; ) are also adopted.  ... 
doi:10.1609/aaai.v35i5.16553 fatcat:mhopklmtlvct7b2oscxkrntvbu

Hyper-SAGNN: a self-attention based graph neural network for hypergraphs [article]

Ruochi Zhang, Yuesong Zou, Jian Ma
2019 arXiv   pre-print
Hyper-SAGNN will be useful for graph representation learning to uncover complex higher-order interactions in different applications.  ...  Here we develop a new self-attention based graph neural network called Hyper-SAGNN applicable to homogeneous and heterogeneous hypergraphs with variable hyperedge sizes.  ...  Conclusion In this work, we have developed a new graph neural network model called Hyper-SAGNN for the representation learning of general hypergraphs.  ... 
arXiv:1911.02613v1 fatcat:ktalcl2hfnhuvkwjydbtmmqou4

An Information-Explainable Random Walk Based Unsupervised Network Representation Learning Framework on Node Classification Tasks

Xin Xu, Yang Lu, Yupeng Zhou, Zhiguo Fu, Yanjie Fu, Minghao Yin
2021 Mathematics  
Network representation learning aims to learn low-dimensional, compressible, and distributed representational vectors of nodes in networks.  ...  This paper proposes an information-explainable random walk based unsupervised network representation learning framework named Probabilistic Accepted Walk (PAW) to obtain network representation from the  ...  How to learn more expressive representational vectors of networks by designing one random walk strategy without hyper-parameters to sample the walking paths with each specific network information individually  ... 
doi:10.3390/math9151767 fatcat:syipmpvosrat3axvnud6yuqjqa

Heterogeneous Graph Matching Networks [article]

Shen Wang, Zhengzhang Chen, Xiao Yu, Ding Li, Jingchao Ni, Lu-An Tang, Jiaping Gui, Zhichun Li, Haifeng Chen, Philip S. Yu
2019 arXiv   pre-print
To address the limitations of existing techniques, we propose MatchGNet, a heterogeneous Graph Matching Network model to learn the graph representation and similarity metric simultaneously based on the  ...  Traditional signature-based malicious program detection algorithms can only detect known malware and are prone to evasion techniques such as binary obfuscation, while behavior-based approaches highly rely  ...  Our path-wise aggregator can learn the attentional weights for different meta-paths automatically.  ... 
arXiv:1910.08074v1 fatcat:hgqew4qbwzfvppiy53vsgeynyu

Improving Hypernymy Detection with an Integrated Path-based and Distributional Method

Vered Shwartz, Yoav Goldberg, Ido Dagan
2016 Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
We suggest an improved path-based algorithm, in which the dependency paths are encoded using a recurrent neural network, that achieves results comparable to distributional methods.  ...  Distributional methods, whose supervised variants are the current best performers, and path-based methods, which received less research attention.  ...  Acknowledgments We would like to thank Omer Levy for his involvement and assistance in the early stage of this project and Enrico Santus for helping us by computing the results of SLQS (Santus et al.,  ... 
doi:10.18653/v1/p16-1226 dblp:conf/acl/ShwartzGD16 fatcat:yljga7p5t5borfg5bneh4s25bi
« Previous Showing results 1 — 15 out of 32,187 results