Linear feature sensibility for output partitioning in ordered neural incremental attribute learning

Ting Wang; Sheng Uei Guan; Jieming Ma; Fangzhou Liu

doi:10.1007/978-3-319-23862-3_37

Linear feature sensibility for output partitioning in ordered neural incremental attribute learning

Ting Wang^*, Sheng Uei Guan, Jieming Ma, Fangzhou Liu

^*Corresponding author for this work

Department of Computing

Research output: Chapter in Book or Report/Conference proceeding › Conference Proceeding › peer-review

Abstract

Feature Ordering is a special training preprocessing for Incremental Attribute Learning (IAL), where features are trained one after another. Since most feature ordering calculation methods, compute feature ordering in one batch, no matter, this study presents a novel approach combining input feature ordered training and output partitioning for IAL to compute feature ordering with considering whether the output of the classification problem is univariate or multivariate. New metric called feature’s Single Sensibility (SS) is proposed to individually calculate features’ discrimination ability for each output. Finally, experimental benchmark results based on neural networks in IAL show that SS is applicable to calculates feature’s discrimination ability. Furthermore, combined output partitioning can also improve further the final classification performance effectively.

Original language	English
Title of host publication	Intelligence Science and Big Data Engineering
Subtitle of host publication	Big Data and Machine Learning Techniques - 5th International Conference, IScIDE 2015, Revised Selected Papers
Editors	Zhi-Hua Zhou, Baochuan Fu, Fuyuan Hu, Zhancheng Zhang, Zhi-Yong Liu, Yanning Zhang, Xiaofei He, Xinbo Gao
Publisher	Springer Verlag
Pages	373-383
Number of pages	11
ISBN (Print)	9783319238616
DOIs	https://doi.org/10.1007/978-3-319-23862-3_37
Publication status	Published - 2015
Event	5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015 - Suzhou, China Duration: 14 Jun 2015 → 16 Jun 2015

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9243
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015
Country/Territory	China
City	Suzhou
Period	14/06/15 → 16/06/15

Keywords

Fisher linear discriminant
Incremental attribute learning
Machine learning
Neural networks
Pattern classification

Access to Document

10.1007/978-3-319-23862-3_37

Cite this

Wang, T., Guan, S. U., Ma, J., & Liu, F. (2015). Linear feature sensibility for output partitioning in ordered neural incremental attribute learning. In Z.-H. Zhou, B. Fu, F. Hu, Z. Zhang, Z.-Y. Liu, Y. Zhang, X. He, & X. Gao (Eds.), Intelligence Science and Big Data Engineering: Big Data and Machine Learning Techniques - 5th International Conference, IScIDE 2015, Revised Selected Papers (pp. 373-383). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9243). Springer Verlag. https://doi.org/10.1007/978-3-319-23862-3_37

Wang, Ting ; Guan, Sheng Uei ; Ma, Jieming et al. / Linear feature sensibility for output partitioning in ordered neural incremental attribute learning. Intelligence Science and Big Data Engineering: Big Data and Machine Learning Techniques - 5th International Conference, IScIDE 2015, Revised Selected Papers. editor / Zhi-Hua Zhou ; Baochuan Fu ; Fuyuan Hu ; Zhancheng Zhang ; Zhi-Yong Liu ; Yanning Zhang ; Xiaofei He ; Xinbo Gao. Springer Verlag, 2015. pp. 373-383 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{d611c8a6feb041768700f65cbc2d777e,

title = "Linear feature sensibility for output partitioning in ordered neural incremental attribute learning",

abstract = "Feature Ordering is a special training preprocessing for Incremental Attribute Learning (IAL), where features are trained one after another. Since most feature ordering calculation methods, compute feature ordering in one batch, no matter, this study presents a novel approach combining input feature ordered training and output partitioning for IAL to compute feature ordering with considering whether the output of the classification problem is univariate or multivariate. New metric called feature{\textquoteright}s Single Sensibility (SS) is proposed to individually calculate features{\textquoteright} discrimination ability for each output. Finally, experimental benchmark results based on neural networks in IAL show that SS is applicable to calculates feature{\textquoteright}s discrimination ability. Furthermore, combined output partitioning can also improve further the final classification performance effectively.",

keywords = "Fisher linear discriminant, Incremental attribute learning, Machine learning, Neural networks, Pattern classification",

author = "Ting Wang and Guan, {Sheng Uei} and Jieming Ma and Fangzhou Liu",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; 5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015 ; Conference date: 14-06-2015 Through 16-06-2015",

year = "2015",

doi = "10.1007/978-3-319-23862-3_37",

language = "English",

isbn = "9783319238616",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "373--383",

editor = "Zhi-Hua Zhou and Baochuan Fu and Fuyuan Hu and Zhancheng Zhang and Zhi-Yong Liu and Yanning Zhang and Xiaofei He and Xinbo Gao",

booktitle = "Intelligence Science and Big Data Engineering",

}

Wang, T, Guan, SU , Ma, J & Liu, F 2015, Linear feature sensibility for output partitioning in ordered neural incremental attribute learning. in Z-H Zhou, B Fu, F Hu, Z Zhang, Z-Y Liu, Y Zhang, X He & X Gao (eds), Intelligence Science and Big Data Engineering: Big Data and Machine Learning Techniques - 5th International Conference, IScIDE 2015, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9243, Springer Verlag, pp. 373-383, 5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015, Suzhou, China, 14/06/15. https://doi.org/10.1007/978-3-319-23862-3_37

Linear feature sensibility for output partitioning in ordered neural incremental attribute learning. / Wang, Ting; Guan, Sheng Uei ; Ma, Jieming et al.
Intelligence Science and Big Data Engineering: Big Data and Machine Learning Techniques - 5th International Conference, IScIDE 2015, Revised Selected Papers. ed. / Zhi-Hua Zhou; Baochuan Fu; Fuyuan Hu; Zhancheng Zhang; Zhi-Yong Liu; Yanning Zhang; Xiaofei He; Xinbo Gao. Springer Verlag, 2015. p. 373-383 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9243).

Research output: Chapter in Book or Report/Conference proceeding › Conference Proceeding › peer-review

TY - GEN

T1 - Linear feature sensibility for output partitioning in ordered neural incremental attribute learning

AU - Wang, Ting

AU - Guan, Sheng Uei

AU - Ma, Jieming

AU - Liu, Fangzhou

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2015.

PY - 2015

Y1 - 2015

N2 - Feature Ordering is a special training preprocessing for Incremental Attribute Learning (IAL), where features are trained one after another. Since most feature ordering calculation methods, compute feature ordering in one batch, no matter, this study presents a novel approach combining input feature ordered training and output partitioning for IAL to compute feature ordering with considering whether the output of the classification problem is univariate or multivariate. New metric called feature’s Single Sensibility (SS) is proposed to individually calculate features’ discrimination ability for each output. Finally, experimental benchmark results based on neural networks in IAL show that SS is applicable to calculates feature’s discrimination ability. Furthermore, combined output partitioning can also improve further the final classification performance effectively.

AB - Feature Ordering is a special training preprocessing for Incremental Attribute Learning (IAL), where features are trained one after another. Since most feature ordering calculation methods, compute feature ordering in one batch, no matter, this study presents a novel approach combining input feature ordered training and output partitioning for IAL to compute feature ordering with considering whether the output of the classification problem is univariate or multivariate. New metric called feature’s Single Sensibility (SS) is proposed to individually calculate features’ discrimination ability for each output. Finally, experimental benchmark results based on neural networks in IAL show that SS is applicable to calculates feature’s discrimination ability. Furthermore, combined output partitioning can also improve further the final classification performance effectively.

KW - Fisher linear discriminant

KW - Incremental attribute learning

KW - Machine learning

KW - Neural networks

KW - Pattern classification

UR - http://www.scopus.com/inward/record.url?scp=84945920643&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-23862-3_37

DO - 10.1007/978-3-319-23862-3_37

M3 - Conference Proceeding

AN - SCOPUS:84945920643

SN - 9783319238616

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 373

EP - 383

BT - Intelligence Science and Big Data Engineering

A2 - Zhou, Zhi-Hua

A2 - Fu, Baochuan

A2 - Hu, Fuyuan

A2 - Zhang, Zhancheng

A2 - Liu, Zhi-Yong

A2 - Zhang, Yanning

A2 - He, Xiaofei

A2 - Gao, Xinbo

PB - Springer Verlag

T2 - 5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015

Y2 - 14 June 2015 through 16 June 2015

ER -

Wang T, Guan SU , Ma J, Liu F. Linear feature sensibility for output partitioning in ordered neural incremental attribute learning. In Zhou ZH, Fu B, Hu F, Zhang Z, Liu ZY, Zhang Y, He X, Gao X, editors, Intelligence Science and Big Data Engineering: Big Data and Machine Learning Techniques - 5th International Conference, IScIDE 2015, Revised Selected Papers. Springer Verlag. 2015. p. 373-383. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-23862-3_37

Linear feature sensibility for output partitioning in ordered neural incremental attribute learning

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Cite this