Multi-source transfer learning for non-stationary environments

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Standard

Multi-source transfer learning for non-stationary environments. / Du, Honghui; Minku, Leandro L.; Zhou, Huiyu .

2019 International Joint Conference on Neural Networks (IJCNN). IEEE Computer Society, 2019. (Neural Networks (IJCNN), International Joint Conference on).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Harvard

Du, H, Minku, LL & Zhou, H 2019, Multi-source transfer learning for non-stationary environments. in 2019 International Joint Conference on Neural Networks (IJCNN). Neural Networks (IJCNN), International Joint Conference on, IEEE Computer Society, International Joint Conference on Neural Networks (IJCNN 2019), Budapest, Hungary, 14/07/19. https://doi.org/10.1109/IJCNN.2019.8852024

APA

Du, H., Minku, L. L., & Zhou, H. (2019). Multi-source transfer learning for non-stationary environments. In 2019 International Joint Conference on Neural Networks (IJCNN) (Neural Networks (IJCNN), International Joint Conference on). IEEE Computer Society. https://doi.org/10.1109/IJCNN.2019.8852024

Vancouver

Du H, Minku LL, Zhou H. Multi-source transfer learning for non-stationary environments. In 2019 International Joint Conference on Neural Networks (IJCNN). IEEE Computer Society. 2019. (Neural Networks (IJCNN), International Joint Conference on). https://doi.org/10.1109/IJCNN.2019.8852024

Author

Du, Honghui ; Minku, Leandro L. ; Zhou, Huiyu . / Multi-source transfer learning for non-stationary environments. 2019 International Joint Conference on Neural Networks (IJCNN). IEEE Computer Society, 2019. (Neural Networks (IJCNN), International Joint Conference on).

Bibtex

@inproceedings{bc757e683bdf4ffe9ffb784a910870d7,
title = "Multi-source transfer learning for non-stationary environments",
abstract = "In data stream mining, predictive models typically suffer drops in predictive performance due to concept drift. As enough data representing the new concept must be collected for the new concept to be well learnt, the predictive performance of existing models usually takes some time to recover from concept drift. To speed up recovery from concept drift and improve predictive performance in data stream mining, this work proposes a novel approach called Multi-sourcE onLine TrAnsfer learning for Non-statIonary Environments (Melanie). Melanie is the first approach able to transfer knowledge between multiple data streaming sources in non-stationary environments. It creates several sub-classifiers to learn different aspects from different source and target concepts over time. The sub-classifiers that match the current target concept well are identified, and used to compose an ensemble for predicting examples from the target concept. We evaluate Melanie on several synthetic data streams containing different types of concept drift and on real world data streams. The results indicate that Melanie can deal with a variety drifts and improve predictive performance over existing data stream learning algorithms by making use of multiple sources.",
keywords = "concept drift, non-stationary environment, multi-sources, transfer learning",
author = "Honghui Du and Minku, {Leandro L.} and Huiyu Zhou",
year = "2019",
month = sep,
day = "30",
doi = "10.1109/IJCNN.2019.8852024",
language = "English",
isbn = "9781728119861",
series = "Neural Networks (IJCNN), International Joint Conference on",
publisher = "IEEE Computer Society",
booktitle = "2019 International Joint Conference on Neural Networks (IJCNN)",
address = "United States",
note = " International Joint Conference on Neural Networks (IJCNN 2019) ; Conference date: 14-07-2019 Through 19-07-2019",

}

RIS

TY - GEN

T1 - Multi-source transfer learning for non-stationary environments

AU - Du, Honghui

AU - Minku, Leandro L.

AU - Zhou, Huiyu

PY - 2019/9/30

Y1 - 2019/9/30

N2 - In data stream mining, predictive models typically suffer drops in predictive performance due to concept drift. As enough data representing the new concept must be collected for the new concept to be well learnt, the predictive performance of existing models usually takes some time to recover from concept drift. To speed up recovery from concept drift and improve predictive performance in data stream mining, this work proposes a novel approach called Multi-sourcE onLine TrAnsfer learning for Non-statIonary Environments (Melanie). Melanie is the first approach able to transfer knowledge between multiple data streaming sources in non-stationary environments. It creates several sub-classifiers to learn different aspects from different source and target concepts over time. The sub-classifiers that match the current target concept well are identified, and used to compose an ensemble for predicting examples from the target concept. We evaluate Melanie on several synthetic data streams containing different types of concept drift and on real world data streams. The results indicate that Melanie can deal with a variety drifts and improve predictive performance over existing data stream learning algorithms by making use of multiple sources.

AB - In data stream mining, predictive models typically suffer drops in predictive performance due to concept drift. As enough data representing the new concept must be collected for the new concept to be well learnt, the predictive performance of existing models usually takes some time to recover from concept drift. To speed up recovery from concept drift and improve predictive performance in data stream mining, this work proposes a novel approach called Multi-sourcE onLine TrAnsfer learning for Non-statIonary Environments (Melanie). Melanie is the first approach able to transfer knowledge between multiple data streaming sources in non-stationary environments. It creates several sub-classifiers to learn different aspects from different source and target concepts over time. The sub-classifiers that match the current target concept well are identified, and used to compose an ensemble for predicting examples from the target concept. We evaluate Melanie on several synthetic data streams containing different types of concept drift and on real world data streams. The results indicate that Melanie can deal with a variety drifts and improve predictive performance over existing data stream learning algorithms by making use of multiple sources.

KW - concept drift

KW - non-stationary environment

KW - multi-sources

KW - transfer learning

UR - http://www.scopus.com/inward/record.url?scp=85073188383&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2019.8852024

DO - 10.1109/IJCNN.2019.8852024

M3 - Conference contribution

SN - 9781728119861

T3 - Neural Networks (IJCNN), International Joint Conference on

BT - 2019 International Joint Conference on Neural Networks (IJCNN)

PB - IEEE Computer Society

T2 - International Joint Conference on Neural Networks (IJCNN 2019)

Y2 - 14 July 2019 through 19 July 2019

ER -