Semi-supervised learning with graphs

By Xiaojin Zhu, John D. Lafferty, R. Rosenfeld. 2005

715 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/3bc4736f9b8512043ed47357a81f26b93a1204b6

How researchers see this paper

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

Citations in the corpus

Similarity-based Classification: Concepts and Algorithms (2009, background): “Lastly, we note that similarity-based learning can be considered a special case of graph-based learning (see, for example, [57]), where the graph is fully-connected.”
Adaptive Hypergraph Learning and its Application in Image Classification (2012, unknown): “Among existing transductive learning methods, graph-based learning [5]–[8], [10], [11], [13], [14], [40], [44], [49] achieves promising performance.”
Seeing Stars: Exploiting Class Relationships for Sentiment Categorization with Respect to Rating Scales (2005, background): “Of course, there are many other related semi-supervised learning algorithms that we would like to try as well; see Zhu (2005) for a survey.”
Learning Context-Sensitive Shape Similarity by Graph Transduction (2010, methodology): “Usually, GFHF is used instead of label propagation, as both methods can achieve the same results [16].”
An Adaptive Semisupervised Feature Analysis for Video Semantic Recognition (2018, background): “Therefore, semisupervised learning for feature selection that aims to learn a discriminative subset of the original features from both labeled and unlabeled data, is of potentially great practical significance [17]–[19].”
Small Data Challenges in Big Data Era: A Survey of Recent Progress on Unsupervised and Semi-Supervised Methods (2019, methodology): “…it can be revealed that the functional gradient over the manifold is closely connected with the Laplace-Beltrami operator Therefore, one can directly calculate the Laplace-Beltrami operator without the approximate graph-based Laplacian that is often used in classic semi-supervised methods [84].”
Storage Fit Learning with Feature Evolvable Streams (2020, methodology): “We focus on manifold regularization which assumes that similar samples should have the same label and has been successfully applied in many practical tasks [ZLR05].”
Video suggestion and discovery for youtube: taking random walks through the view graph (2008, background): “There is a variety of metrics one could choose from (e.g., shortest distance, commute time or electrical resistance, etc.), but most of these are expensive to compute, especially for large graphs.”
Semi-Supervised Learning via Generalized Maximum Entropy (2010, methodology): “On the other hand, graph-based SSL methods (Zhu, 2005) are typically based on the manifold assumption.”
Online domain adaptation of a pre-trained cascade of classifiers (2011, unknown): “In Zhu’s self-training [20], a classifier is first trained using a small set of labeled examples and then the predictions from this trained model are used to re-train the classifier.”
A Survey of Deep Active Learning (2020, background): “s experienced clinicians to complete, and it is clearly impractical to demand that such professionals complete a large amount of medical image labeling. Similar fields also include speech recognition [1, 188], medical imaging [64, 93, 109, 176], recommender systems [2, 26], information extraction [1”
Unifying Graph Convolutional Neural Networks and Label Propagation (2020, methodology): “…on the label side L , node labels are propagated and aggregated along edges in the graph, which is known as Label Propagation Algorithm (LPA) (Zhu et al., 2005; Zhou et al., 2004; Zhang & Lee, 2007; Wang & Zhang, 2008; Karasuyama & Mamitsuka, 2013; Gong et al., 2017; Liu et al., 2019a); On…”
How Complex Is Your Classification Problem? (2018, methodology): “easures, the Gower distance is employed. Two nodes i and jare connected only if dist(i;j) <. This corresponds to the -NN method for building a graph from a dataset in the attribute-value format [Zhu et al., 2005]. As in Morais and Prati [2013] and Garcia et al. [2015], in ECoL the value is set to”