New ensemble methods for evolving data streams

By Albert Bifet, G. Holmes, Bernhard Pfahringer, Richard Kirkby, Ricard Gavaldà. 2009

633 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

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

In Signal Processing

Citations in the corpus

Memory Efficient Experience Replay for Streaming Learning (2018, neutral): “[37], [38], [39], [56], [63] or ensemble methods [7], [9], [27], [44], [46], [51], [54], [61], [70], [72].”
Lifelong Machine Learning with Deep Streaming Linear Discriminant Analysis (2019, methodology): “Another widely used method is ensembling multiple clas-siﬁers [5, 6, 48].”
Ensemble learning: A survey (2018, methodology): “(cid:129) Computational cost : The complexity cost for training the ensemble and the required prediction time for new instances may be considered as important criteria in assessing ensemble models, especially in real-time systems (Bifet et al., 2009).”
Extremely Fast Decision Tree (2018, background): “The de facto standard for learning decision trees from streaming data is Hoeffding Tree (HT) [11], which is used as a base for many state-of-the-art drift learners [3, 6, 8, 10, 16, 18, 24].”
A comprehensive survey of anomaly detection techniques for high dimensional big data (2020, methodology): “The framework and approach of Bifet et al. [127] was similar to that proposed by Narasimhamurthy and Kuncheva [128], which accommodates STAGGER and moving hyperplane generation strategies.”
MOA: Massive Online Analysis, a Framework for Stream Classification and Clustering (2010, neutral): “It uses the sigmoid function, as an elegant and practical solution [9, 10].”
MOA: Massive Online Analysis (2010, methodology): “It uses the sigmoid function, as an elegant and practical solution (Bifet et al., 2009a,b).”
Adaptive random forests for evolving data stream classification (2017, methodology): “Ensemble learners are often preferred when learning from evolving data streams, since they are able to achieve high learning performance, without much optimization, and have the advantageous characteristic of being flexible as new learners can be selectively added, updated, reset or removed [32,14,1”
Reacting to Different Types of Concept Drift: The Accuracy Updated Ensemble Algorithm (2014, methodology): “The third real benchmark data are the Poker dataset [39], which consists of 1 000 000 examples describing the suits and ranks of a hand of five playing cards.”
Leveraging Bagging for Evolving Data Streams (2010, methodology): “We use the Kappa statistic κ [6] to show how using Leveraging Bagging with λ = 6, we increase the diversity of the ensemble.”
A survey on concept drift adaptation (2014, methodology): “Within the framework, it is possible to define the probability that instances of the stream belong to the new concept after the drift using the sigmoid function, as an elegant and practical solution [Bifet et al. 2009].”
Ensemble approaches for regression: A survey (2012, methodology): “We believe that ensemble pruning will regain some importance in the application of ensemble methods to data streams [Wang et al. 2003; Bifet et al. 2009].”
Consistency of Online Random Forests (2013, background): “Many of the online techniques incorporate features designed for this problem (Gama et al., 2005; Abdulsalam, 2008; Saffari et al., 2009; Bifet et al., 2009; 2012).”
A survey on instance selection for active learning (2012, background): “Most existing work on streaming data platform rely on building accurate ensemble models [83].”
Ensemble learning for data stream analysis: A survey (2017, methodology): “The same ADWIN method may also be integrated with online bagging ensemble – see ADWINBagging [18] .”
Scalable and efficient multi-label classification for evolving data streams (2012, methodology): “Synthetic data streams can be generated in addition to using real-world data sets, including concept drift, as it is presented in Bifet et al. (2009).”
On expressiveness and uncertainty awareness in rule-based classification for data streams (2017, positive): “[9,13,15] also use this dataset in their empirical evaluations among other datasets.”
Online Incremental Machine Learning Platform for Big Data-Driven Smart Traffic Management (2019, background): “Existing literature reports a number of supervised machine learning algorithms that detect drifts and adapt to new concepts [3]–[6].”