Dynamic Ensemble Selection Based on Rough Set Reduction and Cluster Matching

Ying-Chun CHEN; Ou LI; Yu SUN

doi:10.1587/transcom.2017EBP3441

Dynamic Ensemble Selection Based on Rough Set Reduction and Cluster Matching

Ying-Chun CHEN, Ou LI, Yu SUN

Full Text Views

0

Cite this

Summary :

Ensemble learning is widely used in the field of sensor network monitoring and target identification. To improve the generalization ability and classification precision of ensemble learning, we first propose an approximate attribute reduction algorithm based on rough sets in this paper. The reduction algorithm uses mutual information to measure attribute importance and introduces a correction coefficient and an approximation parameter. Based on a random sampling strategy, we use the approximate attribute reduction algorithm to implement the multi-modal sample space perturbation. To further reduce the ensemble size and realize a dynamic subset of base classifiers that best matches the test sample, we define a similarity parameter between the test samples and training sample sets that takes the similarity and number of the training samples into consideration. We then propose a k-means clustering-based dynamic ensemble selection algorithm. Simulations show that the multi-modal perturbation method effectively selects important attributes and reduces the influence of noise on the classification results. The classification precision and runtime of experiments demonstrate the effectiveness of the proposed dynamic ensemble selection algorithm.

Publication: IEICE TRANSACTIONS on Communications Vol.E101-B No.10 pp.2196-2202

Publication Date: 2018/10/01

Publicized: 2018/04/11

Online ISSN: 1745-1345

DOI: 10.1587/transcom.2017EBP3441

Type of Manuscript: PAPER

Category: Fundamental Theories for Communications

Authors

Ying-Chun CHEN
  National Digital Switching System Engineering & Technological Research and Development Center(NDSC)
Ou LI
  National Digital Switching System Engineering & Technological Research and Development Center(NDSC)
Yu SUN
  National Digital Switching System Engineering & Technological Research and Development Center(NDSC)

Keyword

mutual information, rough set, attribute reduction, dynamic ensemble selection, k-means

Cite this

Copy

Ying-Chun CHEN, Ou LI, Yu SUN, "Dynamic Ensemble Selection Based on Rough Set Reduction and Cluster Matching" in IEICE TRANSACTIONS on Communications, vol. E101-B, no. 10, pp. 2196-2202, October 2018, doi: 10.1587/transcom.2017EBP3441.
Abstract: Ensemble learning is widely used in the field of sensor network monitoring and target identification. To improve the generalization ability and classification precision of ensemble learning, we first propose an approximate attribute reduction algorithm based on rough sets in this paper. The reduction algorithm uses mutual information to measure attribute importance and introduces a correction coefficient and an approximation parameter. Based on a random sampling strategy, we use the approximate attribute reduction algorithm to implement the multi-modal sample space perturbation. To further reduce the ensemble size and realize a dynamic subset of base classifiers that best matches the test sample, we define a similarity parameter between the test samples and training sample sets that takes the similarity and number of the training samples into consideration. We then propose a k-means clustering-based dynamic ensemble selection algorithm. Simulations show that the multi-modal perturbation method effectively selects important attributes and reduces the influence of noise on the classification results. The classification precision and runtime of experiments demonstrate the effectiveness of the proposed dynamic ensemble selection algorithm.
URL: https://global.ieice.org/en_transactions/communications/10.1587/transcom.2017EBP3441/_p

Copy

@ARTICLE{e101-b_10_2196,
author={Ying-Chun CHEN, Ou LI, Yu SUN, },
journal={IEICE TRANSACTIONS on Communications},
title={Dynamic Ensemble Selection Based on Rough Set Reduction and Cluster Matching},
year={2018},
volume={E101-B},
number={10},
pages={2196-2202},
abstract={Ensemble learning is widely used in the field of sensor network monitoring and target identification. To improve the generalization ability and classification precision of ensemble learning, we first propose an approximate attribute reduction algorithm based on rough sets in this paper. The reduction algorithm uses mutual information to measure attribute importance and introduces a correction coefficient and an approximation parameter. Based on a random sampling strategy, we use the approximate attribute reduction algorithm to implement the multi-modal sample space perturbation. To further reduce the ensemble size and realize a dynamic subset of base classifiers that best matches the test sample, we define a similarity parameter between the test samples and training sample sets that takes the similarity and number of the training samples into consideration. We then propose a k-means clustering-based dynamic ensemble selection algorithm. Simulations show that the multi-modal perturbation method effectively selects important attributes and reduces the influence of noise on the classification results. The classification precision and runtime of experiments demonstrate the effectiveness of the proposed dynamic ensemble selection algorithm.},
keywords={},
doi={10.1587/transcom.2017EBP3441},
ISSN={1745-1345},
month={October},}

Copy

TY - JOUR
TI - Dynamic Ensemble Selection Based on Rough Set Reduction and Cluster Matching
T2 - IEICE TRANSACTIONS on Communications
SP - 2196
EP - 2202
AU - Ying-Chun CHEN
AU - Ou LI
AU - Yu SUN
PY - 2018
DO - 10.1587/transcom.2017EBP3441
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E101-B
IS - 10
JA - IEICE TRANSACTIONS on Communications
Y1 - October 2018
AB - Ensemble learning is widely used in the field of sensor network monitoring and target identification. To improve the generalization ability and classification precision of ensemble learning, we first propose an approximate attribute reduction algorithm based on rough sets in this paper. The reduction algorithm uses mutual information to measure attribute importance and introduces a correction coefficient and an approximation parameter. Based on a random sampling strategy, we use the approximate attribute reduction algorithm to implement the multi-modal sample space perturbation. To further reduce the ensemble size and realize a dynamic subset of base classifiers that best matches the test sample, we define a similarity parameter between the test samples and training sample sets that takes the similarity and number of the training samples into consideration. We then propose a k-means clustering-based dynamic ensemble selection algorithm. Simulations show that the multi-modal perturbation method effectively selects important attributes and reduces the influence of noise on the classification results. The classification precision and runtime of experiments demonstrate the effectiveness of the proposed dynamic ensemble selection algorithm.
ER -