Combining active learning and semi-supervised learning using Gaussian fields and harmonic functions

By Xiaojin Zhu, Zoubin Ghahramani, John D. Lafferty. 2003

3,471 citations recorded by the source corpus.

Subfields: Representation & Self-Supervised Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Representation & Self-Supervised Learning

Citations in the corpus

Semi-Supervised Classification with Graph Convolutional Networks (2016, methodology): “Prominent examples for graph Laplacian regularization include label propagation (Zhu et al., 2003), manifold regularization (Belkin et al.”
Graph Attention Networks (2017, methodology): “…metrics already reported in Hamilton et al. (2017) for (Belkin et al., 2006) 59.5% 60.1% 70.7% SemiEmb (Weston et al., 2012) 59.0% 59.6% 71.7% LP (Zhu et al., 2003) 68.0% 45.3% 63.0% DeepWalk (Perozzi et al., 2014) 67.2% 43.2% 65.3% ICA (Lu & Getoor, 2003) 75.1% 69.1% 73.9% Planetoid (Yang et…”
Self-Training With Noisy Student Improves ImageNet Classification (2019, methodology): “Finally, frameworks in semi-supervised learning also include graph-based methods [100, 87, 92, 42], methods that make use of latent variables as target variables [41, 53, 93] and methods based on low-density separation [26, 69, 19], which might provide complementary benefits to our method.”
Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs (2016, background): “The last application we consider is learning dense intrinsic correspondence between collections of 3D shapes represented as discrete manifolds.”
MixMatch: A Holistic Approach to Semi-Supervised Learning (2019, neutral): “, “transductive” models [14, 22, 21], graph-based methods [49, 4, 29], generative modeling [3, 27, 41, 9, 17, 23, 38, 34, 42], etc.”
Semi-supervised Learning with Deep Generative Models (2014, methodology): “It would be desirable to have a single principled loss function similar to (Blum et al., 2004) or (Zhu et al., 2003).”
Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning (2016, background): “However, with around 1000 images per category, it barely contains enough training samples to prevent the ConvNet from overfitting [7].”
Seeing through the Human Reporting Bias: Visual Classifiers from Noisy Human-Centric Labels (2015, background): “SSL approaches are often computationally impractical [56, 58] or make strong independence assumptions [13] that do not hold in human-centric annotations.”
Smart Reply: Automated Response Suggestion for Email (2016, background): “Other graph-based semi-supervised learning techniques have been explored in the past for more traditional classification problems [27, 5].”
A Survey on Data Collection for Machine Learning: A Big Data - AI Integration Perspective (2018, background): “[102] proposed a semi-supervised learning based on a Gaussian random field model where the unlabeled and labeled examples form a weighted graph.”
Deep Graph Infomax (2018, methodology): “…50 runs of training (followed by logistic regression), and reuse the metrics already reported in Kipf & Welling (2016a) for the performance of DeepWalk and GCN, as well as Label Propagation (LP) (Zhu et al., 2003) and Planetoid (Yang et al., 2016)—a representative supervised random walk method.”
Masked Label Prediction: Unified Massage Passing Model for Semi-Supervised Classification (2020, background): “, 2019) and the Label Propagation Algorithms (LPAs) (Zhu, 2005; Zhu et al., 2003; Zhang & Lee, 2007; Wang & Zhang, 2007; Karasuyama & Mamitsuka, 2013; Gong et al., 2016; Liu et al., 2019).”
Simplifying Graph Convolutional Networks (2019, background): “Remarkably, Label Propagation (Zhu et al., 2003) makes predictions by spreading label information from labeled nodes to their neighbors, which is similar to feature propagation in SGC.”
Good Semi-supervised Learning That Requires a Bad GAN (2017, methodology): “Traditional graph-based methods [2, 26] were extended to deep neural networks [22, 23, 8], which involves applying convolutional neural networks [10] and feature learning techniques to graphs so that the underlying manifold structure can be exploited.”