Filters








15,854 Hits in 5.3 sec

Semi-supervised Learning for Convolutional Neural Networks via Online Graph Construction [article]

Sheng-Yi Bai, Sebastian Agethen, Ting-Hsuan Chao, Winston Hsu
2016 arXiv   pre-print
In this work, we revisit graph-based semi-supervised learning algorithms and propose an online graph construction technique which suits deep convolutional neural network better.  ...  We consider an EM-like algorithm for semi-supervised learning on deep neural networks: In forward pass, the graph is constructed based on the network output, and the graph is then used for loss calculation  ...  To model the structural density, graph is thus constructed to measure the proximity (similarity) between data points in graph-based semi-supervised learning.  ... 
arXiv:1511.06104v2 fatcat:6z7ocup6kja25c3kohl47yfxha

Semi-supervised Feature Selection via Spectral Analysis [chapter]

Zheng Zhao, Huan Liu
2007 Proceedings of the 2007 SIAM International Conference on Data Mining  
We present a semi-supervised feature selection algorithm based on spectral analysis.  ...  The paucity of labeled instances provides insufficient information about the structure of the target concept, and can cause supervised feature selection algorithms to fail.  ...  The task of learning from mixed labeled and unlabeled data is of semi-supervised learning [2] .  ... 
doi:10.1137/1.9781611972771.75 dblp:conf/sdm/ZhaoL07 fatcat:kl4hxy2dhfhzpcssnd4ykajbiu

Analysis of label noise in graph-based semi-supervised learning

Bruno Klaus de Aquino Afonso, Lilian Berton
2020 Proceedings of the 35th Annual ACM Symposium on Applied Computing  
Semi-supervised learning (SSL) alleviates that by making strong assumptions about the relation between the labels and the input data distribution.  ...  Our work aims to perform an extensive empirical evaluation of existing graph-based semi-supervised algorithms, like Gaussian Fields and Harmonic Functions, Local and Global Consistency, Laplacian Eigenmaps  ...  RELATED WORK Graph-based methods have been a staple of semi-supervised learning for some time. Many of the graph-based semi-supervised methods may be formulated as a convex optimization method.  ... 
doi:10.1145/3341105.3374013 dblp:conf/sac/AfonsoB20 fatcat:mtfoihu7pze2lgejps5jugrxeu

Semi-supervised Data Representation via Affinity Graph Learning [article]

Weiya Ren
2015 arXiv   pre-print
A new semi-supervised learning framework is proposed by combing manifold regularization and data representation methods such as Non negative matrix factorization and sparse coding.  ...  The proposed framework forms the Laplacian regularizer through learning the affinity graph.  ...  A variety of graph-based semi-supervised learning (GSSL) [1, 5, 8, 9, 10, 13] have recently become popular due to their high accuracy and computational efficiency.  ... 
arXiv:1502.03879v1 fatcat:badyxyo4pbf27mgyvt7tsyuqha

Iterative Graph Self-Distillation [article]

Hanlin Zhang, Shuai Lin, Weiyang Liu, Pan Zhou, Jian Tang, Xiaodan Liang, Eric P. Xing
2021 arXiv   pre-print
As a natural extension, we also apply IGSD to semi-supervised scenarios by jointly regularizing the network with both supervised and unsupervised contrastive loss.  ...  Empirically, we achieve significant and consistent performance gain on various graph datasets in both unsupervised and semi-supervised settings, which well validates the superiority of IGSD.  ...  [55] improves performance in GNN-based semi-supervised node classification via edge prediction.  ... 
arXiv:2010.12609v2 fatcat:h5csmfxatbg4jcliukikimsgnm

KGNN: Harnessing Kernel-based Networks for Semi-supervised Graph Classification [article]

Wei Ju, Junwei Yang, Meng Qu, Weiping Song, Jianhao Shen, Ming Zhang
2022 arXiv   pre-print
The GNN-based network performs classification through learning graph representations to implicitly capture the similarity between query graphs and labeled graphs, while the kernel-based network uses graph  ...  This paper studies semi-supervised graph classification, which is an important problem with various applications in social network analysis and bioinformatics.  ...  The goal of semi-supervised graph classification is to learn a label distribution 𝑝 (y 𝑈 |G, y 𝐿 ), which can assign labels to unlabeled graphs G 𝑈 .  ... 
arXiv:2205.10550v1 fatcat:shd52l46wbgsvptdtz3zqriy5a

Logistic label propagation

Takumi Kobayashi, Kenji Watanabe, Nobuyuki Otsu
2012 Pattern Recognition Letters  
In this paper, we propose a novel method for semi-supervised learning, called logistic label propagation (LLP).  ...  To cope with unlabeled samples as well as labeled ones in the semi-supervised learning framework, the logistic functions are learnt by using similarities between samples in a manner similar to label propagation  ...  the semi-supervised learning.  ... 
doi:10.1016/j.patrec.2011.12.005 fatcat:xp476dvmbvfwtpge4ebp2qw47e

Learning Classifiers on a Partially Labeled Data Manifold

Qiuhua Liu, Xuejun Liao, Lawrence Carin
2007 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07  
We present an algorithm for learning parametric classifiers on a partially labeled data manifold, based on a graph representation of the manifold.  ...  The unlabeled data are utilized by basing classifier learning on neighborhoods, formed via Markov random walks.  ...  both labeled and unlabeled data, thus enforcing semi-supervised learning.  ... 
doi:10.1109/icassp.2007.366312 dblp:conf/icassp/LiuLC07 fatcat:kzohrm7ohbgvnb2viel6w25xoe

Semi-Supervised Classification with Graph Convolutional Networks [article]

Thomas N. Kipf, Max Welling
2017 arXiv   pre-print
We present a scalable approach for semi-supervised learning on graph-structured data that is based on an efficient variant of convolutional neural networks which operate directly on graphs.  ...  We motivate the choice of our convolutional architecture via a localized first-order approximation of spectral graph convolutions.  ...  This problem can be framed as graph-based semi-supervised learning, where label information is smoothed over the graph via some form of explicit graph-based regularization (Zhu et al., 2003; Zhou et al  ... 
arXiv:1609.02907v4 fatcat:pt7vcc7b5vecflnsu4xdigvceq

Shoestring: Graph-Based Semi-Supervised Learning with Severely Limited Labeled Data [article]

Wanyu Lin, Zhaolin Gao, Baochun Li
2020 arXiv   pre-print
More specifically, we address the problem of graph-based semi-supervised learning in the presence of severely limited labeled samples, and propose a new framework, called Shoestring, that improves the  ...  Graph-based semi-supervised learning has been shown to be one of the most effective approaches for classification tasks from a wide range of domains, such as image classification and text classification  ...  Revisiting Graph-based Semi-Supervised Learning We do not attempt to provide a comprehensive literature review on graph-based semi-supervised learning.  ... 
arXiv:1910.12976v2 fatcat:bc2uikyhgnb6hifwzdw2m6psny

Label Efficient Semi-Supervised Learning via Graph Filtering [article]

Qimai Li, Xiao-Ming Wu, Han Liu, Xiaotong Zhang, Zhichao Guan
2019 arXiv   pre-print
In this paper, we address label efficient semi-supervised learning from a graph filtering perspective.  ...  Graph-based methods have been demonstrated as one of the most effective approaches for semi-supervised learning, as they can exploit the connectivity patterns between labeled and unlabeled data samples  ...  Revisit and Extend Label Propagation Label propagation (LP) [63, 61, 5] is arguably the most popular method for graph-based semi-supervised learning.  ... 
arXiv:1901.09993v3 fatcat:oqwhmisb4bbstgak5y4gqgruyu

A Semi-Supervised Assessor of Neural Architectures [article]

Yehui Tang, Yunhe Wang, Yixing Xu, Hanting Chen, Chunjing Xu, Boxin Shi, Chao Xu, Qi Tian, Chang Xu
2020 arXiv   pre-print
A graph convolutional neural network is introduced to predict the performance of architectures based on the learned representations and their relation modeled by the graph.  ...  In contrast with classical performance predictor optimized in a fully supervised way, this paper suggests a semi-supervised assessor of neural architectures.  ...  Though the supervised loss is only applied on labeled architectures, the unlabeled architectures also participate in the performance prediction of the labeled architectures via the relation graph G, and  ... 
arXiv:2005.06821v1 fatcat:6qmboccmmveohimkedbmbqqxbm

Bayesian Semi-supervised Learning with Graph Gaussian Processes [article]

Yin Cheng Ng, Nicolo Colombo, Ricardo Silva
2018 arXiv   pre-print
We propose a data-efficient Gaussian process-based Bayesian approach to the semi-supervised learning problem on graphs.  ...  in active learning experiments where labels are scarce.  ...  Many of the successful graph-based semi-supervised learning models are based on graph Laplacian regularization or learning embeddings of the nodes.  ... 
arXiv:1809.04379v3 fatcat:ynsxjoq4k5gdbk7ziesvm77leu

Adversarial Semi-supervised Learning for Corporate Credit Ratings [article]

Bojing Feng, Wenfang Xue
2021 arXiv   pre-print
Then in the second phase, adversarial semi-supervised learning is applied uniting labeled data and pseudo-labeled data.  ...  Specifically, we consider the problem of adversarial semi-supervised learning (ASSL) for corporate credit rating which has been rarely researched before.  ...  In this work, we explore how to extract knowledge from unlabeled data via adversarial semi-supervised learning.  ... 
arXiv:2104.02479v2 fatcat:hgnqbyc5krabtnrfnw2kvlirme

Graph construction andb-matching for semi-supervised learning

Tony Jebara, Jun Wang, Shih-Fu Chang
2009 Proceedings of the 26th Annual International Conference on Machine Learning - ICML '09  
Graph based semi-supervised learning (SSL) methods play an increasingly important role in practical machine learning systems.  ...  This article provides an empirical study of leading semi-supervised methods under a wide range of graph construction algorithms.  ...  based semi-supervised learning performs well.  ... 
doi:10.1145/1553374.1553432 dblp:conf/icml/JebaraWC09 fatcat:newolkrrtbfofcdbum5iexvm34
« Previous Showing results 1 — 15 out of 15,854 results