Filters








66,968 Hits in 4.4 sec

Handling Missing Data with Graph Representation Learning [article]

Jiaxuan You, Xiaobai Ma, Daisy Yi Ding, Mykel Kochenderfer, Jure Leskovec
2020 arXiv   pre-print
Machine learning with missing data has been approached in two different ways, including feature imputation where missing feature values are estimated based on observed values, and label prediction where  ...  GRAPE tackles the missing data problem using a graph representation, where the observations and features are viewed as two types of nodes in a bipartite graph, and the observed feature values as edges.  ...  In contrast, GRAPE handles the missing feature entries naturally with the graph representation without any additional heuristics.  ... 
arXiv:2010.16418v1 fatcat:i5uxtsxfabhajlfvp4fyke53lq

TSI-GNN: Extending Graph Neural Networks to Handle Missing Data in Temporal Settings

David Gordon, Panayiotis Petousis, Henry Zheng, Davina Zamanzadeh, Alex A.T. Bui
2021 Frontiers in Big Data  
We present a novel approach for imputing missing data that incorporates temporal information into bipartite graphs through an extension of graph representation learning.  ...  To the best of our knowledge, this is the first effort to use a joint bipartite graph approach that captures sequence information to handle missing data.  ...  In this work, we introduce temporal setting imputation using graph neural networks (TSI-GNN), which extends graph representation learning to handle missing data in temporal settings.  ... 
doi:10.3389/fdata.2021.693869 pmid:34604740 pmcid:PMC8480427 fatcat:7ovshd4q3jeyhl5z6kj2ijg7su

Siamese Attribute-missing Graph Auto-encoder [article]

Wenxuan Tu, Sihang Zhou, Yue Liu, Xinwang Liu
2021 arXiv   pre-print
Graph representation learning (GRL) on attribute-missing graphs, which is a common yet challenging problem, has recently attracted considerable attention.  ...  In this paper, based on the idea of introducing intimate information interaction between the two information sources, we propose our Siamese Attribute-missing Graph Auto-encoder (SAGA).  ...  Handling Missing Data with Graph Representation Learning. In NeurIPS. You, Y.; Chen, T.; Sui, Y.; Chen, T.; Wang, Z.; and Shen, Y. 2020b. Graph Contrastive Learning with Augmentations. In NeurIPS.  ... 
arXiv:2112.04842v1 fatcat:2hcdamwg5zf2diopgvp2whjdhi

Wasserstein Graph Neural Networks for Graphs with Missing Attributes [article]

Zhixian Chen, Tengfei Ma, Yangqiu Song, Yang Wang
2022 arXiv   pre-print
Graph neural networks have been demonstrated power in graph representation learning while their performance is affected by the completeness of graph information.  ...  In addition, we find WGNN suitable to recover missing values and adapt them to tackle matrix completion problems with graphs of users and items.  ...  Machine learning with missing data. To handle missing data, most machine learning methods rely on data imputation.  ... 
arXiv:2102.03450v2 fatcat:v7upfmhjsrfjljzylf57fx5lnq

Unified Graph-Based Missing Label Propagation Method for Multilabel Text Classification

Adil Yaseen Taha, Sabrina Tiun, Abdul Hadi Abd Rahman, Masri Ayob, Ali Sabah Abdulameer
2022 Symmetry  
In order to address the incomplete or missing label problem, this study proposes two methods: an aggregated feature and label graph-based missing label handling method (GB-AS), and a unified graph-based  ...  A high-order label correlation is learned from the incomplete training data and applied to supplement the missing label matrix, which guides the creation of multilabel classification models.  ...  with feature similarity weighting (GB-FS); • Graph-based missing label handling with label similarity weighting (GB-LS).  ... 
doi:10.3390/sym14020286 fatcat:gdq73ccxzzbfld5zq7zt2okqke

Graph Augmentation Learning [article]

Shuo Yu, Huafei Huang, Minh N. Dao, Feng Xia
2022 arXiv   pre-print
Graph Augmentation Learning (GAL) provides outstanding solutions for graph learning in handling incomplete data, noise data, etc.  ...  The aggregation mechanism of augmentation strategies and graph learning models are also discussed by different application scenarios, i.e., data-specific, model-specific, and hybrid scenarios.  ...  2021] , but cases in graph representation learning are still missing.  ... 
arXiv:2203.09020v1 fatcat:72esohvrxbdzdlnahaa27pftqm

Data Pre-Processing and Customized Onto-Graph Construction for Knowledge Extraction in Healthcare Domain of Semantic Web

2019 VOLUME-8 ISSUE-10, AUGUST 2019, REGULAR ISSUE  
Missing values were randomly induced into the Pima Indian Diabetic dataset with the missing ratio of 1%, 3% and 5% for each attribute up to 50% of the attributes in the original diabetic dataset.  ...  Further, importance of knowledge graph and various ontological representation types are discussed in short as construction of .owl file is the first step towards automation in semantic web.  ...  handling missing data is one among the most important data pre-processing steps.  ... 
doi:10.35940/ijitee.k1423.0981119 fatcat:wbfpy4uz6rgyrlhlty6chhx5r4

Learning Representations of Missing Data for Predicting Patient Outcomes [article]

Brandon Malone, Alberto Garcia-Duran, Mathias Niepert
2018 arXiv   pre-print
Patients are often missing data relative to each other; the data comes in a variety of modalities, such as multivariate time series, free text, and categorical demographic information; important relationships  ...  In this work, we propose a novel approach to address these first three challenges using a representation learning scheme based on message passing.  ...  Learning an explicit representation of missing data within the graph-based learning framework has the advantage of propagating representations of missing data throughout the graph such that these representations  ... 
arXiv:1811.04752v1 fatcat:3gnr53fm2fhlblaswbrm4anrri

Medication Recommendation and Lab Test Imputation via Graph Convolutional Networks [article]

Chengsheng Mao, Liang Yao, Yuan Luo
2021 arXiv   pre-print
Then we model the graph to learn a distributed representation for each entity in the graph based on graph convolutional networks (GCN).  ...  In our system, we integrate the complex relations between multiple types of medical entities with their inherent features in a heterogeneous graph.  ...  Graph Representation Learning on EHR Graph representation learning tries to learn a vector representation for part or the whole of a graph for certain graph learning tasks, e.g., learning node representation  ... 
arXiv:1904.00326v2 fatcat:vspwubeshba2nbfcut5lkjdqsi

Error-Robust Multi-View Clustering: Progress, Challenges and Opportunities [article]

Mehrnaz Najafi and Lifang He and Philip S. Yu
2021 arXiv   pre-print
With recent advances in data collection from multiple sources, multi-view data has received significant attention. In multi-view data, each view represents a different perspective of data.  ...  Existing error-robust multi-view clustering approaches with explicit error removal formulation can be structured into five broad research categories - sparsity norm based approaches, graph based methods  ...  Besides, one of the recent deep learning work invokes GAN to generate missing data instances in incomplete views while learning the shared representation for clustering .  ... 
arXiv:2105.03058v1 fatcat:dwgokdsehbfhxebdfazt2xg7hu

Graph embedded incomplete multi-view clustering method with proximity relation estimation

Runze Chen, Jie Wen, Xiaoyue Chen, Yong Xu
2020 The Journal of Engineering  
However, in reality, this is often not the case and the real data are always with incomplete views, which lead to the failure of the conventional methods.  ...  of samples in different views to complete their affinity graphs.  ...  By minimising the disagreement of kernel matrices constructed from low-dimensional representation of data in diverse views, the method learns a common compact representation of original data.  ... 
doi:10.1049/joe.2019.1164 fatcat:jj57gaplvfazrab3knedjgj34q

Spatial-Temporal Graph Attention Networks: A Deep Learning Approach for Traffic Forecasting

Chenhan Zhang, James J. Q. Yu, Yi Liu
2019 IEEE Access  
Nonetheless, traditional methods show their limitation in coping with complexity and high nonlinearity of traffic data as well as learning spatial-temporal dependencies.  ...  In this paper, we propose a novel deep learning framework, Spatial-Temporal Graph Attention Networks (ST-GAT).  ...  TABLE 3 . 3 Accuracy of ST-GAT with measurement noise and missing data.  ... 
doi:10.1109/access.2019.2953888 fatcat:h3mjwe3765bophjanjexw6gs24

Incomplete Multi-view Clustering via Graph Regularized Matrix Factorization [article]

Jie Wen, Zheng Zhang, Yong Xu, Zuofeng Zhong
2018 arXiv   pre-print
Compared with the conventional graph embedding methods, the proposed method does not introduce any extra regularization term and corresponding penalty parameter to preserve the local structure of data,  ...  Moreover, the proposed method can be viewed as a unified framework for multi-view learning since it can handle both incomplete and complete multi-view clustering and classification tasks.  ...  MIC jointly learns the latent representation of each view and the consensus representation by utilizing the weighted NMF algorithm, in which the missing views are constrained with the small weight even  ... 
arXiv:1809.05998v1 fatcat:476imxvbfvgu3ojzgup2repmxy

Incomplete Multi-view Clustering via Graph Regularized Matrix Factorization [chapter]

Jie Wen, Zheng Zhang, Yong Xu, Zuofeng Zhong
2019 Lecture Notes in Computer Science  
Compared with the conventional graph embedding methods, the proposed method does not introduce any extra regularization term and corresponding penalty parameter to preserve the local structure of data,  ...  Moreover, the proposed method can be viewed as a unified framework for multi-view learning since it can handle both incomplete and complete multi-view clustering and classification tasks.  ...  MIC jointly learns the latent representation of each view and the consensus representation by utilizing the weighted NMF algorithm, in which the missing views are constrained with the small weight even  ... 
doi:10.1007/978-3-030-11018-5_47 fatcat:jpdovxdnvrbavneli52k2ylmpu

A partitioning approach to scaling anomaly detection in graph streams

William Eberle, Lawrence Holder
2014 2014 IEEE International Conference on Big Data (Big Data)  
Due to potentially complex relationships among heterogeneous data sets, recent research efforts have involved the representation of this type of complex data as a graph.  ...  To address this issue, we propose a novel approach called Pattern Learning and Anomaly Detection on Streams, or PLADS, that is not only scalable to real-world data that is streaming, but also maintains  ...  In addition, they have not dealt with the scalability issues associated with "big data" when attempting to learn patterns and anomalies in data represented as a graph.  ... 
doi:10.1109/bigdata.2014.7004367 dblp:conf/bigdataconf/EberleH14 fatcat:uuozvojevngillq4v7nc3yfpwm
« Previous Showing results 1 — 15 out of 66,968 results