Classifier Ensembles for Changing Environments

By Ludmila I. Kuncheva. 2004

381 citations recorded by the source corpus.

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

From amateurs to connoisseurs: modeling the evolution of user expertise through online reviews (2013, background): “Some early works that deal with temporal dynamics do so in terms of concept drift [23, 37, 39].”
A framework for generating data to simulate changing environments (2007, background): “Bespoke online classification models [5, 6, 32], as well as classifier ensembles [21,29] have been proposed for changing environments [22].”
Learning in Nonstationary Environments: A Survey (2015, background): “20 IEEE CompuTaTIonal InTEllIgEnCE magazInE | novEmbEr 2015 and offer some distinct advantages: (i) they tend to be more accurate than single classifier-based systems due to reduction in the variance of the error; (ii) they have the flexibility to easily incorporate new data into a classification mo”
Incremental Learning of Concept Drift in Nonstationary Environments (2011, background): “Drift speed describes the displacement rate in P t (ω | x ) from one time step to the next, P t + 1 (ω | x ) .”
A Survey on Ensemble Learning for Data Stream Classification (2017, methodology): “We refer readers to chapter 8 of Kuncheva’s book [Kuncheva 2004b] and other works [Kuncheva and Whitaker 2003; Banﬁeld et al. 2005; Zhou 2012] for an extensive discussion of diversity measures for ensemble learners.”
The CART decision tree for mining data streams (2014, background): “Additionally, the concept drift may occur [8,10,17,20,22,34], which means that the concept of data evolve in time.”
Adaptive Machine Learning for Credit Card Fraud Detection (2015, background): “State-of-the-art 46 It becomes critical then to set the rate of forgetting, defined by the window size, in order to match the rate of change in the distribution [165].”
Applying lazy learning algorithms to tackle concept drift in spam filtering (2007, methodology): “Building on the analysis presented in Kuncheva (2004), the techniques for using ensemble to handle concept drift fall into two groups (Delany, Cuningham, & Tsymbal, 2005): (i) dynamic combiners where the base classiﬁers are trained in advance and the concept drift is tracked by changing the…”
Covariate Shift Adaptation by Importance Weighted Cross Validation (2007, background): “, 2005) or EPSRC (Kuncheva, 2006), Common benchmark data sets can be used to evaluate the experimental performance of proposed and related methods.”
Ensemble based systems in decision making (2006, background): “The long list includes composite classifier systems [1], mixture of experts [4], [5], stacked generalization [6], consensus aggregation [7], combination of multiple classifiers [8]–[12], change-glasses approach to classifier selection [13], dynamic classifier selection [12], classifier fusion [14]–[”
On-line Boosting and Vision (2006, background): “This (current) classifier is updated whenever a new sample arrives [63].”
Classifier Ensembles for Detecting Concept Change in Streaming Data: Overview and Perspectives (2008, methodology): “Included in this group are most ensemble methods for changing environments: Winnow variants [24, 28], AdaBoost variants [5, 30], “replace-the-loser” approaches [21,22,34,35,39].”
Ensemble learning for data stream analysis: A survey (2017, methodology): “45, 65, 116], and some have been written several years ago [109, 111, 60].”
Reacting to Different Types of Concept Drift: The Accuracy Updated Ensemble Algorithm (2014, methodology): “In recent years, several approaches that cope with concept drift have been proposed, including mainly: sliding window approaches, new online algorithms, special detection tech iques, and adaptive ensembles; for their systematic revie ws see [1], [7].”