Incorporating Diversity in Active Learning with Support Vector Machines

By Klaus Brinker. International Conference on Machine Learning, 2003

572 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/0ca966cb390b442b10cb76aa3fddee6b613f4f0f

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

Semi-automated screening of biomedical citations for systematic reviews (2010, methodology): “The heuristic used to decide whether or not the space of minority examples has been adequately explored is a function of the diversity over the encountered minority examples thus far, as measured by the average angle between each pair in kernel-space [28].”
What Do I Annotate Next? An Empirical Study of Active Learning for Action Localization (2018, background): “To alleviate this problem, researchers often consider active learning in batch-mode [3].”
Predicting sample size required for classification performance (2012, background): “Detailed information about the three algorithms can be found in appendix 2 (see additional file 2) and in literature [10,35,42].”
A Survey of Active Learning Algorithms for Supervised Remote Sensing Image Classification (2011, background): “In applicative scenarios, diversity among samples [42] is highly desirable.”
Active Learning for Convolutional Neural Networks: A Core-Set Approach (2017, background): “, 2009), and geometric distance to decision boundaries (Tong & Koller, 2001; Brinker, 2003).”
Active and Semisupervised Learning for the Classification of Remote Sensing Images (2014, methodology): “Brinker introduced an SVM-based batch approach, which selects a batch of samples that minimizes the margin while maximizing their diversity [41].”
A machine learning approach to sentiment analysis in multilingual Web texts (2009, background): “2004; SaarTsechansky and Provost 2004; Osugi 2005) or a set of examples (Brinker 2003; Xu et al. 2003; Nguyen and Smeulders 2004) is selected, labeled by a human and added to the training set in order to retrain the classifier until a stop condition is reached.”
Using text mining for study identification in systematic reviews: a systematic review of current approaches (2015, unknown): “They do this drawing on the work of Brinker who has developed methods for incorporating diversity in active learning by evaluating the stability of a measure of similarity between ‘included’ citations between iterations [75].”
Entity Linking with a Knowledge Base: Issues, Techniques, and Solutions (2015, methodology): “[72] used an instance selection strategy (similar to active learning [131], [132]) to select a more balanced and informative subset”
Multi-Class Active Learning by Uncertainty Sampling with Diversity Maximization (2015, background): “In Brinker (2003), diversity constraint is combined with SVM for active learning.”
Adversarial Active Learning for Deep Networks: a Margin Based Approach (2018, background): “If, exploiting the geometric distances has been relevant for active learning on SVM (Tong & Koller, 2001; Brinker, 2003), it is not intuitive for CNNs since we do not know beforehand the geometrical shape of their decision boundaries.”
Active Learning Literature Survey (2009, background): “Brinker (2003) considers an approach for SVMs that explicitly incorporates diversity among instances in the batch.”
Batch-Mode Active-Learning Methods for the Interactive Classification of Remote Sensing Images (2011, methodology): “In order to address this shortcoming, in [13] an approach is presented especially designed to construct batches of samples by incorporating a diversity measure that c onsiders the angles between the induced classification hyperplanes (more details on this ap proach are given in the next section).”
Variational Adversarial Active Learning (2019, background): “In non-Bayesian classical active learning approaches, uncertainty heuristics such as distance from the decision boundary, highest entropy, and expected risk minimization have been widely investigated [3, 50, 54].”
From Theories to Queries (2011, background): “, by ranking the most informative instances, and then clustering those that are most highly ranked (Brinker, 2003; Kim et al., 2006; Xu et al., 2007).”
Cost-Effective Active Learning for Deep Image Classification (2017, background): “Moreover, the diversity of the selected instance over the certain category has been taken into consideration in [4] as well.”
Can Language Models Encode Perceptual Structure Without Grounding? A Case Study in Color (2021, background): “…might be beneﬁcial for robust representation learning, in correspondence with the idea of sample diversity in the active learning literature (Brinker, 2003; Yang et al., 2015). pos-ent ’s lack of signiﬁcance, on the other hand, indicates that the degree of speciﬁcation offered by the POS…”
A Survey of Deep Active Learning (2020, background): “se a hybrid query strategy in a batch query, taking into account both the information volume and diversity of samples in either an explicit or implicit manner. Early Batch Mode Active Learning (BMAL) [20, 77, 111, 171, 173] algorithm performance is often excessively reliant on the measurement of sim”