Learning by Transduction

By A. Gammerman, Vladimir G Vovk, Vladimir Naumovich Vapnik. 1998

We describe a method for predicting a classification of an object given classifications of the objects in the training set, assuming that the pairs object/classification are generated by an i.i.d. process from a continuous probability distribution. Our method is a modification of Vapnik's support-vector machine; its main novelty is that it gives not only the prediction itself but also a practicable measure of the evidence found in support of that prediction. We also describe a procedure for assigning degrees of confidence to predictions made by the support vector machine. Some experimental results are presented, and possible extensions of the algorithms are discussed.

Subfields: Probabilistic & Bayesian Methods, Statistical Learning Theory

How researchers see this paper

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

In Probabilistic & Bayesian Methods

In Statistical Learning Theory

Citations in the corpus

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.”
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).”
Discriminative Improvements to Distributional Sentence Similarity (2013, background): “Matrix factorization on both training and (unlabeled) test data can be viewed as a form of transductive learning (Gammerman et al., 1998), where we assume access to unlabeled test set instances.”
Denoising Pre-Training and Data Augmentation Strategies for Enhanced RDF Verbalization with Transformers (2020, unknown): “In the context of the WebNLG challenge, we considered a transduction strategy (Gammerman et al., 1998).”
Cross-Domain Object Recognition Via Input-Output Kernel Analysis (2013, methodology): “Transductive SVM [2], [24], Domain Adaptive SVM and cross-domain SVM [25], and other variants of SVMs are also explored in domain adaptation problems by defining a new decision function incorporating data from bot h the target and source domains.”
Data mining in the Life Sciences with Random Forest: a walk in the park or lost in the jungle? (2012, methodology): “Additionally, RF has been incorporated in a transductive confidence machine [95], a framework that allows the prediction of classifiers to be complemented with a confidence value that can be set by the user prior to classification [103].”
Inductive Confidence Machines for Regression (2002, background): “We will actually prove the stronger assertion that (3) is true if the randomization is done only over the calibration set and the new example.”
On the Equivalence between Positional Node Embeddings and Structural Graph Representations (2019, background): “In their original definitions, transductive learning (Gammerman et al., 1998), (Zhu et al.”
Transduction with Confidence and Credibility (1999, methodology): “, 1998; Gammerman, 1997] it provides confidence and credibility values (the predictive performance however, of the algorithm described in [Gammerman et al., 1998] is poor; this is probably explained by the “distortion phenomenon” : see [Gammerman et al.”
Machine learning framework for early MRI-based Alzheimer’s conversion prediction in MCI subjects (2014, unknown): “For a clear explanation of the differences between transductive and inductive machine learning algorithms, we refer to Gammerman et al. (1998) and relation between semi-supervised and transductive learning is discussed in detail by Chapelle et al. (2006).”
Open set face recognition using transduction (2005, methodology): “The p p p p-value construction used here has been proposed by Gammerman et al. [11] and Proedrou et al. [28].”
Federated Semi-Supervised Learning with Inter-Client Consistency (2020, background): “While there exist numerous work on SSL, such as transductive models [18], graph-based approaches [19], and generative modeling [20], we mainly discuss consistency regularization and entropy-minimization approaches.”
Tent: Fully Test-Time Adaptation by Entropy Minimization (2021, background): “Transduction (Gammerman et al., 1998; Joachims, 1999; Zhou et al., 2004) jointly optimizes on train and test to better fit specific test instances.”
Test-Time Training with Self-Supervision for Generalization under Distribution Shifts (2019, background): “Lastly, we share the philosophy of transductive learning (Vapnik, 2013; Gammerman et al., 1998), but have little in common with their classical algorithms; recent work by Tripuraneni & Mackey (2019) theoretically explores this for linear prediction, in the context of debiasing the LASSO estimator.”
Ridge Regression Learning Algorithm in Dual Variables (1998, methodology): “[3] to obtain a measure of confidence for predictions output by our algorithms.”
Learning Transferrable Representations for Unsupervised Domain Adaptation (2016, background): “Transductive learning: In the transductive learning [10], the model has access to unlabelled test samples during training.”