Diverse ensembles for active learning

By Prem Melville, Raymond J. Mooney. International Conference on Machine Learning, 2004

401 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/7e71466744358ec0a2f503228f017531442b5090

How researchers see this paper

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

In Artificial Intelligence

Citations in the corpus

Active Learning Methods for Remote Sensing Image Classification (2009, methodology): “In [39], results obtained by query-by-boosting and query-by-bagging are compared on several batch data”
Hyperspectral Image Classification—Traditional to Deep Models: A Survey for Future Prospects (2021, methodology): “The committee of different classifiers can either be built by using ensemble learning algorithms like Bagging and Boosting [235] or by changing the model”
Ensemble deep learning: A review (2021, methodology): “In [154, 155], the author explores using an ensemble method that builds a diverse committee.”
Outlier Analysis (2013, methodology): “A variety of such criteria based on ensemble learning are discussed in [331].”
Active Learning Literature Survey (2009, background): “In these domains, active feature acquisition (Zheng and Padmanabhan, 2002; Melville et al., 2004) seeks to alleviate these problems by allowing the learner to request more complete feature information.”
Dual Strategy Active Learning (2007, background): “Our future plan is also to extend this work to ensemble methods that involve more than two strategies, maximizing ensemble diversity [15,14,10].”
A Survey of Active Learning Algorithms for Supervised Remote Sensing Image Classification (2011, methodology): “In [26], results obtained by query-by-boosting and query-by-bagging are compared using several batch datasets showing excellent performance of the heuristics proposed.”
Committee-Based Active Learning for Surrogate-Assisted Particle Swarm Optimization of Expensive Problems (2017, background): “Among them, query by committee (QBC) [38] is a simple yet very effective active learning approach [39], which can”
Active Learning (2012, methodology): “Melville and Mooney (2004) proposed a different ensemble-based method that explicitly encourages diversity among the committee members. Muslea et al. (2000) constructed a committee of hypotheses by partitioning the feature space into conditionally independent committee-specific subsets, to be used i”
Creating diversity in ensembles using artificial data (2005, background): “In addition to improving classification accuracy in a traditional supervised setting, recent work has shown that Decorate can be very effective for active learning (Melville and Mooney, 2004).”
Outlier detection by active learning (2006, background): “(See, for example, [ 12 ].) More precisely, we define the margin as follows.”
Active Multi-task Learning via Bandits (2015, methodology): “According to the query strategies used, there are three categories of active learning techniques: (1) uncertainty sampling [21, 31], which focuses on selecting the instance that the classifier is most uncertain about; (2) query by committee [16, 24], in which the most informative instance is the one”
Variational Adversarial Active Learning (2019, methodology): “Using an ensemble of models to represent uncertainty was proposed by [30, 55], but [38] showed that using ensembles does not always yield high diversity in predictions which results in sampling redundant instances.”