Reliability-guided fuzzy classifier ensemble

Tianhua Chen, Pan Su, Changjing Shang, Qiang Shen

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

1 Citation (Scopus)

Abstract

Classifier ensembles form an important approach to improving classification performance. As such, there have been different proposals made in the literature that provide a range of means to construct and aggregate classifier ensembles. However, the resulting systems may contain unreliable members with false or biased judgements in the ensemble. The removal of unreliable members is necessary to optimise the overall performance of such systems. Smaller ensembles involving reduced ensemble members also helps relax the requirement of computational memory, thereby strengthening the system's run-time efficiency. To differentiate the potential contributions of different ensemble members while reducing the adverse impact of any unreliable judgement upon the system, a nearest neighbour-based reliability measure is incorporated into the process of classifier ensemble selection. In particular, reliabilities of selected ensemble members are perceived as a stress function, from which argument-dependent weights are heuristically generated for final aggregated decision. Experimental investigations are carried out, demonstrating the efficacy of the proposed approach, where fuzzy classifiers are utilised as base members of the emerging ensemble.
LanguageEnglish
Title of host publication2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)
PublisherIEEE
ISBN (Electronic)9781509060344
ISBN (Print)9781509060351
DOIs
Publication statusPublished - 24 Aug 2017
Externally publishedYes
EventIEEE International Conference on Fuzzy Systems - Royal Continental Hotel, Naples, Italy
Duration: 9 Jul 201712 Jul 2017
http://www.fuzzieee2017.org/ (Link to Conference Website)

Conference

ConferenceIEEE International Conference on Fuzzy Systems
Abbreviated titleFUZZ-IEEE 2017
CountryItaly
CityNaples
Period9/07/1712/07/17
Internet address

Fingerprint

Classifiers
Data storage equipment

Cite this

Chen, T., Su, P., Shang, C., & Shen, Q. (2017). Reliability-guided fuzzy classifier ensemble. In 2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) IEEE. https://doi.org/10.1109/FUZZ-IEEE.2017.8015407
Chen, Tianhua ; Su, Pan ; Shang, Changjing ; Shen, Qiang. / Reliability-guided fuzzy classifier ensemble. 2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE). IEEE, 2017.
@inproceedings{581505a7569b485990a5d34a6fb36303,
title = "Reliability-guided fuzzy classifier ensemble",
abstract = "Classifier ensembles form an important approach to improving classification performance. As such, there have been different proposals made in the literature that provide a range of means to construct and aggregate classifier ensembles. However, the resulting systems may contain unreliable members with false or biased judgements in the ensemble. The removal of unreliable members is necessary to optimise the overall performance of such systems. Smaller ensembles involving reduced ensemble members also helps relax the requirement of computational memory, thereby strengthening the system's run-time efficiency. To differentiate the potential contributions of different ensemble members while reducing the adverse impact of any unreliable judgement upon the system, a nearest neighbour-based reliability measure is incorporated into the process of classifier ensemble selection. In particular, reliabilities of selected ensemble members are perceived as a stress function, from which argument-dependent weights are heuristically generated for final aggregated decision. Experimental investigations are carried out, demonstrating the efficacy of the proposed approach, where fuzzy classifiers are utilised as base members of the emerging ensemble.",
keywords = "Reliability, Open wireless architecture, Training, Atmospheric measurements, Particle measurements, Buildings, Weight measurement",
author = "Tianhua Chen and Pan Su and Changjing Shang and Qiang Shen",
year = "2017",
month = "8",
day = "24",
doi = "10.1109/FUZZ-IEEE.2017.8015407",
language = "English",
isbn = "9781509060351",
booktitle = "2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)",
publisher = "IEEE",

}

Chen, T, Su, P, Shang, C & Shen, Q 2017, Reliability-guided fuzzy classifier ensemble. in 2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE). IEEE, IEEE International Conference on Fuzzy Systems, Naples, Italy, 9/07/17. https://doi.org/10.1109/FUZZ-IEEE.2017.8015407

Reliability-guided fuzzy classifier ensemble. / Chen, Tianhua; Su, Pan ; Shang, Changjing; Shen, Qiang.

2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE). IEEE, 2017.

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

TY - GEN

T1 - Reliability-guided fuzzy classifier ensemble

AU - Chen, Tianhua

AU - Su, Pan

AU - Shang, Changjing

AU - Shen, Qiang

PY - 2017/8/24

Y1 - 2017/8/24

N2 - Classifier ensembles form an important approach to improving classification performance. As such, there have been different proposals made in the literature that provide a range of means to construct and aggregate classifier ensembles. However, the resulting systems may contain unreliable members with false or biased judgements in the ensemble. The removal of unreliable members is necessary to optimise the overall performance of such systems. Smaller ensembles involving reduced ensemble members also helps relax the requirement of computational memory, thereby strengthening the system's run-time efficiency. To differentiate the potential contributions of different ensemble members while reducing the adverse impact of any unreliable judgement upon the system, a nearest neighbour-based reliability measure is incorporated into the process of classifier ensemble selection. In particular, reliabilities of selected ensemble members are perceived as a stress function, from which argument-dependent weights are heuristically generated for final aggregated decision. Experimental investigations are carried out, demonstrating the efficacy of the proposed approach, where fuzzy classifiers are utilised as base members of the emerging ensemble.

AB - Classifier ensembles form an important approach to improving classification performance. As such, there have been different proposals made in the literature that provide a range of means to construct and aggregate classifier ensembles. However, the resulting systems may contain unreliable members with false or biased judgements in the ensemble. The removal of unreliable members is necessary to optimise the overall performance of such systems. Smaller ensembles involving reduced ensemble members also helps relax the requirement of computational memory, thereby strengthening the system's run-time efficiency. To differentiate the potential contributions of different ensemble members while reducing the adverse impact of any unreliable judgement upon the system, a nearest neighbour-based reliability measure is incorporated into the process of classifier ensemble selection. In particular, reliabilities of selected ensemble members are perceived as a stress function, from which argument-dependent weights are heuristically generated for final aggregated decision. Experimental investigations are carried out, demonstrating the efficacy of the proposed approach, where fuzzy classifiers are utilised as base members of the emerging ensemble.

KW - Reliability

KW - Open wireless architecture

KW - Training

KW - Atmospheric measurements

KW - Particle measurements

KW - Buildings

KW - Weight measurement

U2 - 10.1109/FUZZ-IEEE.2017.8015407

DO - 10.1109/FUZZ-IEEE.2017.8015407

M3 - Conference contribution

SN - 9781509060351

BT - 2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)

PB - IEEE

ER -

Chen T, Su P, Shang C, Shen Q. Reliability-guided fuzzy classifier ensemble. In 2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE). IEEE. 2017 https://doi.org/10.1109/FUZZ-IEEE.2017.8015407