Early Drift Detection Method

By Manuel Baena-Garc, J. Avila, Albert Bifet, Ricard Gavald, R. Morales-Bueno. 2005

913 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/2747577a61c70bc3874380130615e15aff76339e

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

The Impact of Diversity on Online Ensemble Learning in the Presence of Concept Drift (2010, background): “However, additional procedures should be adopted in order to properly converge to a new concept, for example, by creating a new ensemble after a drift detection [12], [13], [14], [15].”
Continual Learning in Practice (2019, methodology): “There have been some attempts to tackle this, such as Early Drift Detection Method (EDDM) [31], but this method requires waiting for a minimum of 30 classiﬁcation errors before calculating the monitoring statistic at each decision point, which may be too high for many real applications.”
Learning under Concept Drift: A Review (2019, methodology): “Similar implementations have been adopted and applied in the Learning with Local Drift Detection (LLDD) [25], Early Drift Detection Method (EDDM) [26], Heoffding’s inequality based Drift Detection Method (HDDM) [23], Fuzzy Windowing Drift Detection Method (FW-DDM) [5], Dynamic Extreme Learning Machi”
Reacting to Different Types of Concept Drift: The Accuracy Updated Ensemble Algorithm (2014, methodology): “They also seem to confirm that ensemble approaches that use batch classifiers, such as AWE, may suffer accuracy drops after sudden concept drifts [11], while drift detecto rs are less accurate on gradually drifting streams [24].”
Learning in Nonstationary Environments: A Survey (2015, methodology): “[56] by relying on the analysis of the distance between two classification errors (i.”
Learning under Concept Drift: an Overview (2010, background): “e of drift also referred as gradual includes more than two sources, however, the dierence between the sources is very small, thus the drift is noticed only when looking at a longer time period (e.g. [148, 37, 5]). We refer to the former type of gradual drift as gradual and the latter type of drift a”
Analyzing concept drift and shift from sample data (2018, methodology): “…has been a very substantial effort investigating methods for detecting concept drift (Widmer and Kubat 1996; Kifer et al. 2004; Gama et al. 2004; Baena-Garcıa et al. 2006; Nishida and Yamauchi 2007; Dries and Rückert 2009; Gama and Rodrigues 2009; Žliobaite 2010; Hoens et al. 2012;…”
Advances in Computational Intelligence (2003, unknown): “These algorithms maintain a learning model which is reset upon change detection [141, 24, 289].”