Filters








528,326 Hits in 3.4 sec

Graph Transfer Learning via Adversarial Domain Adaptation with Graph Convolution [article]

Quanyu Dai, Xiao-Ming Wu, Jiaren Xiao, Xiao Shen, Dan Wang
2022 arXiv   pre-print
To tackle this problem, we propose a novel graph transfer learning framework AdaGCN by leveraging the techniques of adversarial domain adaptation and graph convolution.  ...  Existing methods for single network learning cannot solve this problem due to the domain shift across networks.  ...  The idea is to learn class discriminative node representations via graph convolutional networks and learn domain invariant node representations via adversarial learning.  ... 
arXiv:1909.01541v4 fatcat:43rwyp3v3vel3jnho3oyyqaicu

Detecting Malicious Domains via Graph Inference [chapter]

Pratyusa K. Manadhata, Sandeep Yadav, Prasad Rao, William Horne
2014 Lecture Notes in Computer Science  
: 0.99 • 19.7K known good domains: 0.01 • Unknown hosts and domains: 0.5 Benign Malicious Benign 0.51 0.49 Malicious 0.49 0.51 Detection details Low degree false positives Unknown domain  ...  Belief propagation algorithm [P82, YFW01] Marginal probability estimation in graphs • NP-complete Belief propagation is fast and approximate • Iterative message passing Message passing Message(i → j)  ... 
doi:10.1007/978-3-319-11203-9_1 fatcat:5ymcv3vwlvai7ezyudd3uuliya

Detecting Malicious Domains via Graph Inference

Pratyusa Manadhata, Sandeep Yadav, Prasad Rao, William Horne
2014 Proceedings of the 2014 Workshop on Artificial Intelligent and Security Workshop - AISec '14  
: 0.99 • 19.7K known good domains: 0.01 • Unknown hosts and domains: 0.5 Benign Malicious Benign 0.51 0.49 Malicious 0.49 0.51 Detection details Low degree false positives Unknown domain  ...  Belief propagation algorithm [P82, YFW01] Marginal probability estimation in graphs • NP-complete Belief propagation is fast and approximate • Iterative message passing Message passing Message(i → j)  ... 
doi:10.1145/2666652.2666659 dblp:conf/ccs/ManadhataYRH14 fatcat:tytwkqyrg5hjbm6bulqyjjckra

Shape-Biased Domain Generalization via Shock Graph Embeddings [article]

Maruthi Narayanan, Vickram Rajendran, Benjamin Kimia
2021 arXiv   pre-print
in domain generalization.  ...  The resulting graph and its descriptor is a complete representation of contour content and is classified using recent Graph Neural Network (GNN) methods.  ...  We show that this representation of shape information via a shock graph and using GNN to train on them leads to excellent classification performance, even though the appearance is not taken into account  ... 
arXiv:2109.05671v1 fatcat:hza4tljxcbcvhn4zzpzxmda2b4

A simple coding for cross-domain matching with dimension reduction via spectral graph embedding [article]

Hidetoshi Shimodaira
2015 arXiv   pre-print
The cross-domain matching is solved by applying the single-domain version of spectral graph embedding to these augmented vectors of all the domains.  ...  This formulation of cross-domain matching is regarded as an extension of the spectral graph embedding to multi-domain setting, and it includes several multivariate analysis methods of statistics such as  ...  A brief review of the spectral graph embedding The spectral graph theory Before discussing the cross-domain matching, here we review the spectral graph theory (Chung, 1997) .  ... 
arXiv:1412.8380v2 fatcat:5nlazlyeqfe6favnifxrmv5xsy

Domain Adaptation in Physical Systems via Graph Kernel

Haoran Li, Hanghang Tong, Yang Weng
2022 Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining  
Specifically, Domain Adaptation (DA) seeks domain-invariant features to boost the model performance in the target domain.  ...  The spatial structures, temporal trends, measurement similarity, and label information together determine the similarity of two graphs, guiding the DA to find domain-invariant features.  ...  In this paper, we show the integration can be conveniently implemented via a specially-designed graph kernel.  ... 
doi:10.1145/3534678.3539380 fatcat:ctpclqvsanh2ppyvqbvsvkw3l4

Cross-Domain Detection via Graph-Induced Prototype Alignment

Minghao Xu, Hang Wang, Bingbing Ni, Qi Tian, Wenjun Zhang
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
To mitigate these problems, we propose a Graph-induced Prototype Alignment (GPA) framework to seek for category-level domain alignment via elaborate prototype representations.  ...  Applying the knowledge of an object detector trained on a specific domain directly onto a new domain is risky, as the gap between two domains can severely degrade model's performance.  ...  In specific, domain adaptation is realized via aligning two domains' prototypes, in which the critical information of each instance is aggregated via graph-based message propagation, and the multi-modal  ... 
doi:10.1109/cvpr42600.2020.01237 dblp:conf/cvpr/XuWNTZ20 fatcat:2j2xtuqnovgwdmd3up7vgdsuni

Cross-domain Detection via Graph-induced Prototype Alignment [article]

Minghao Xu, Hang Wang, Bingbing Ni, Qi Tian, Wenjun Zhang
2020 arXiv   pre-print
To mitigate these problems, we propose a Graph-induced Prototype Alignment (GPA) framework to seek for category-level domain alignment via elaborate prototype representations.  ...  Applying the knowledge of an object detector trained on a specific domain directly onto a new domain is risky, as the gap between two domains can severely degrade model's performance.  ...  In specific, domain adaptation is realized via aligning two domains' prototypes, in which the critical information of each instance is aggregated via graph-based message propagation, and the multi-modal  ... 
arXiv:2003.12849v1 fatcat:ha3iewm3ovbn7g5s4cpydoavee

Unsupervised Domain Adaptation for Point Cloud Semantic Segmentation via Graph Matching [article]

Yikai Bian, Le Hui, Jianjun Qian, Jin Xie
2022 arXiv   pre-print
Specifically, in order to extract local-level features, we first dynamically construct local feature graphs on both domains and build a memory bank with the graphs from the source domain.  ...  Then, based on the assignment matrix, we can align the feature distributions between the two domains with the graph-based local feature loss.  ...  Then, in order to enrich the graph of the source domain, we construct a feature graph memory bank to store the generated source-domain feature graphs during the training phase.  ... 
arXiv:2208.04510v1 fatcat:o7a4s2mc55d4te6pxj6owf4554

Marginalized Denoising Autoencoder via Graph Regularization for Domain Adaptation [chapter]

Yong Peng, Shen Wang, Bao-Liang Lu
2013 Lecture Notes in Computer Science  
To explicitly preserve the intrinsic structure of data, this paper proposes a marginalized Denoising Autoencoders via graph Regularization (GmSDA) in which the autoencoder based framework can learn more  ...  robust features with the help of newly incorporated graph regularization.  ...  The proposed model, marginalized Denoising Autoencoders via graph Regularization (GmSDA), is introduced in section 3.  ... 
doi:10.1007/978-3-642-42042-9_20 fatcat:ijeyc43g2ncsfhstcnj7w23bgm

Cross-Domain Recommendation via Clustering on Multi-Layer Graphs

Aleksandr Farseev, Ivan Samborskii, Andrey Filchenkov, Tat-Seng Chua
2017 Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval - SIGIR '17  
In addition, we suggest a new approach for automatic construction of inter-network relationship graph based on the data, which eliminates the necessity of having pre-defined domain knowledge.  ...  [52] proposed cross-domain recommender systems, where inter-domain linking was implemented via the so-called "bridge" users (social media users who have accounts on two or more social networks).  ...  The intersource relationship graph is computed automatically from the data and further utilized via novel graph-constrained regularization.  ... 
doi:10.1145/3077136.3080774 dblp:conf/sigir/FarseevSFC17 fatcat:oq6h6njzmffj7lhzub2nsfor6a

PrecoG: an efficient unitary split preconditioner for the transform-domain LMS filter via graph Laplacian regularization [article]

Tamal Batabyal, Daniel S. Weller, Jaideep Kapur, Scott T. Acton
2020 arXiv   pre-print
With the input modeled as a weighted graph that mimics neuronal interactions, PrecoG obtains the desired transform by recursive estimation of the graph Laplacian matrix.  ...  Transform-domain least mean squares (LMS) adaptive filters encompass the class of algorithms where the input data are subjected to a data-independent unitary transform followed by a power normalization  ...  Sparsification is justified in the graph domain, where the sparsifying action leads to O(N ) number of edges and edge weights.  ... 
arXiv:1812.04570v2 fatcat:famo4pnw3zay7oya3l4yuwu2xm

Domain-independent data cleaning via analysis of entity-relationship graph

Dmitri V. Kalashnikov, Sharad Mehrotra
2006 ACM Transactions on Database Systems  
RELDC views the database as a graph of entities that are linked to each other via relationships.  ...  u and v via b, whereas the connection via a is unique to u and v.  ... 
doi:10.1145/1138394.1138401 fatcat:l2encb2zhbbzrm7fwixzinjvjm

Detecting User Community in Sparse Domain via Cross-Graph Pairwise Learning

Zheng Gao, Hongsong Li, Zhuoren Jiang, Xiaozhong Liu
2020 Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval  
While several main cyber domains carrying high-quality graphs, unfortunately, most others can be quite sparse.  ...  However, as users may appear in multiple domains (graphs), their high-quality activities in the main domains can supply community detection in the sparse ones, e.g., user behaviors on Google can help thousands  ...  [6] calibrates domain-specific graph Laplacians into a unified kernel, which detects graph patterns in semi-supervised fashion.  ... 
doi:10.1145/3397271.3401055 dblp:conf/sigir/GaoLJL20 fatcat:uh67hg5rjvbq7m6lix5lqbvgge

Unsupervised Domain Adaptation for Person Re-identification via Heterogeneous Graph Alignment

Minying Zhang, Kai Liu, Yidong Li, Shihui Guo, Hongtao Duan, Yimin Long, Yi Jin
2021 AAAI Conference on Artificial Intelligence  
The proposed domain adaptation framework not only improves model generalization on target domain, but also facilitates mining and integrating the potential discriminative information across different cameras  ...  In this paper, we propose a coarse-tofine heterogeneous graph alignment (HGA) method to find cross-camera person matches by characterizing the unlabeled data as a heterogeneous graph for each camera.  ...  We construct heterogeneous graphs in each camera-specific sub-domain to exploit the potential distribution structure.  ... 
dblp:conf/aaai/ZhangLLGDLJ21 fatcat:6bjsjasfbzftrgiieuvknqdx6y
« Previous Showing results 1 — 15 out of 528,326 results