Transductive Inference for Text Classification using Support Vector Machines

By Thorsten Joachims. 1999

How researchers see this paper

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

In Natural Language Processing

Citations in the corpus

Semi-Supervised Classification with Graph Convolutional Networks (2016, methodology): “We omit TSVM (Joachims, 1999), as it does not scale to the large number of classes in one of our datasets.”
Adversarial Training Methods for Semi-Supervised Text Classification (2016, methodology): “Adversarial and virtual adversarial training resemble some semi-supervised or transductive SVM approaches (Joachims, 1999; Chapelle & Zien, 2005; Collobert et al., 2006; Belkin et al., 2006) in that both families of methods push the decision boundary far from training examples (or in the case of…”
Revisiting Semi-Supervised Learning with Graph Embeddings (2016, methodology): “Among the transductive methods, Planetoid-T achieves the best performance on Cora and Pubmed, while TSVM performs the best on Citeseer.”
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, background): “We compare to a broad range of existing solutions in semi-supervised learning, in particular to classiﬁcation using nearest neighbours (NN), support vector machines on the labelled set (SVM), the transductive SVM (TSVM), the Embedded neural networks Weston et al. (2012), and contractive auto-encoder”
Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning (2016, background): “A notable example is the ’ImageNet’ [5] dataset with 1000 category and more than 1 million training images.”
Meta-Learning for Semi-Supervised Few-Shot Classification (2018, background): “Our algorithm is also related to transductive learning (Vapnik, 1998; Joachims, 1999; Fu et al., 2015), where the base classiﬁer gets reﬁned by seeing the unlabeled examples.”
Semi-supervised learning, causality, and the conditional cluster assumption (2019, methodology): “We compare with T-SVMs [8, 16] with linear and RBF kernels using the q3svm implementation [26], and with label propagation [14, 15] using the implementation in sci-kit learn [27].”
A Baseline for Few-Shot Image Classification (2019, methodology): “Transductive learning: This approach is more efficient at using few labeled data than supervised learning (Joachims, 1999; Zhou et al., 2004; Vapnik, 2013).”
ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring (2019, background): “…research on “transductive” or graph-based semi-supervised learning techniques which leverage the idea that unlabeled datapoints should be assigned the label of a labeled datapoint if they are sufﬁciently similar (Gammerman et al., 1998; Joachims, 2003; 1999; Bengio et al., 2006; Liu et al., 2018).”
Semi-Supervised Self-Training of Object Detection Models (2005, background): “In recent years there has been a substantial amount of work that addresses the problem of incorporating unlabeled data into the training process, [14], [9], [11].”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, methodology): “The domain adaptation SVM (DASVM) [23] exploits within the semi-supervised DA scenario both the transductive SVM [123] and its extension, the progressive transductive SVM [38].”
Smooth Neighbors on Teacher Graphs for Semi-Supervised Learning (2017, background): “Transductive SVMs (TSVM) [17] implement the cluster assumption by keeping unlabeled data far away from the decision boundaries.”
Machine learning framework for early MRI-based Alzheimer’s conversion prediction in MCI subjects (2014, unknown): “The implementation of TSVM was introduced first by Joachims (1999), which assigned a Hinge Loss function L(t) on the labeled samples and Symmetric Hinge Loss function L(|t|) on the unlabeled samples.”
Boosting for transfer learning (2007, methodology): “The constraints designed for TSVM (Joachims, 1999) have also been applied to the basic learners.”