Adaptive random forests for evolving data stream classification

By Heitor Murilo Gomes, Albert Bifet, Jesse Read, Jean Paul Barddal, Fabrício Enembreck, Bernhard Pfahringer, G. Holmes, Talel Abdessalem. 2017

763 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

1. Jie Lu
2. Guangquan Zhang
3. Barbara Hammer
4. Heiko Wersing
5. Anjin Liu
6. Feng Gu
7. Fan Dong
8. João Gama

In Signal Processing

1. Barbara Hammer
2. João Gama
3. Jie Lu
4. Anjin Liu
5. Heiko Wersing
6. Feng Gu
7. Guangquan Zhang
8. Fan Dong

Citations in the corpus

• Memory Efficient Experience Replay for Streaming Learning (2018, neutral): “Many are based on Hoeffding Decision Trees [6], [19], [26], [37], [38], [39], [40], [57], [64] or ensemble methods [8], [10], [28], [45], [47], [52], [55], [62], [71], [73].”
• Learning under Concept Drift: A Review (2019, methodology): “7 KDDCup’99 [106] 494021 41 23 [35], [47], [65], [69], [74], [84], [86], [99], [102]”
• Mitigating Concept Drift via Rejection (2018, methodology): “5 / 11 ARF: Adaptive random forests (ARF) [14] constitute a state-of-the-art ensemble method for learning with drift.”
• Under review as a conference paper at ICLR 2020 many domain adaptation methods (2019, background): “As noticed in the literature Wyner et al. (2017); Gomes et al. (2017), the random forests tend not to propagate severe overfitting error even with a large number of trees.”