Recursive total error rate minimization

Se In Jang; Geok Choo Tan; Kar Ann Toh

doi:10.1109/TENCON.2016.7847954

Recursive total error rate minimization

Se In Jang, Geok Choo Tan, Kar Ann Toh

Department of Electrical and Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In learning algorithm design, a direct optimization of a cost function which fits the goal is naturally desired. Particularly in recursive learning, a direct formulation for total-error-rate (TER) minimization is much desired for online classification applications. However, due to a nonlinear counting step in the classification formulation, an exact solution to minimize TER recursively is yet to be established. In this paper, we propose an exact recursive formulation for TER minimization. Based on empirical evaluations using benchmark data sets, we show that the proposed recursive classification algorithm preserves the performance of the batch mode TER while easing the computational memory load by sample based accumulation.

Original language	English
Title of host publication	Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	38-41
Number of pages	4
ISBN (Electronic)	9781509025961
DOIs	https://doi.org/10.1109/TENCON.2016.7847954
Publication status	Published - 2017 Feb 8
Event	2016 IEEE Region 10 Conference, TENCON 2016 - Singapore, Singapore Duration: 2016 Nov 22 → 2016 Nov 25

Publication series

Name	IEEE Region 10 Annual International Conference, Proceedings/TENCON
ISSN (Print)	2159-3442
ISSN (Electronic)	2159-3450

Other

Other	2016 IEEE Region 10 Conference, TENCON 2016
Country	Singapore
City	Singapore
Period	16/11/22 → 16/11/25

All Science Journal Classification (ASJC) codes

Computer Science Applications
Electrical and Electronic Engineering

Access to Document

10.1109/TENCON.2016.7847954

Fingerprint Dive into the research topics of 'Recursive total error rate minimization'. Together they form a unique fingerprint.

Cite this

Jang, S. I., Tan, G. C., & Toh, K. A. (2017). Recursive total error rate minimization. In Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016 (pp. 38-41). [7847954] (IEEE Region 10 Annual International Conference, Proceedings/TENCON). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/TENCON.2016.7847954

@inproceedings{150f73f55e574a398ee9d622abf36db6,

title = "Recursive total error rate minimization",

abstract = "In learning algorithm design, a direct optimization of a cost function which fits the goal is naturally desired. Particularly in recursive learning, a direct formulation for total-error-rate (TER) minimization is much desired for online classification applications. However, due to a nonlinear counting step in the classification formulation, an exact solution to minimize TER recursively is yet to be established. In this paper, we propose an exact recursive formulation for TER minimization. Based on empirical evaluations using benchmark data sets, we show that the proposed recursive classification algorithm preserves the performance of the batch mode TER while easing the computational memory load by sample based accumulation.",

author = "Jang, {Se In} and Tan, {Geok Choo} and Toh, {Kar Ann}",

year = "2017",

month = feb,

day = "8",

doi = "10.1109/TENCON.2016.7847954",

language = "English",

series = "IEEE Region 10 Annual International Conference, Proceedings/TENCON",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "38--41",

booktitle = "Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016",

address = "United States",

note = "2016 IEEE Region 10 Conference, TENCON 2016 ; Conference date: 22-11-2016 Through 25-11-2016",

}

Jang, SI, Tan, GC & Toh, KA 2017, Recursive total error rate minimization. in Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016., 7847954, IEEE Region 10 Annual International Conference, Proceedings/TENCON, Institute of Electrical and Electronics Engineers Inc., pp. 38-41, 2016 IEEE Region 10 Conference, TENCON 2016, Singapore, Singapore, 16/11/22. https://doi.org/10.1109/TENCON.2016.7847954

Recursive total error rate minimization. / Jang, Se In; Tan, Geok Choo; Toh, Kar Ann.

Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016. Institute of Electrical and Electronics Engineers Inc., 2017. p. 38-41 7847954 (IEEE Region 10 Annual International Conference, Proceedings/TENCON).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Recursive total error rate minimization

AU - Jang, Se In

AU - Tan, Geok Choo

AU - Toh, Kar Ann

PY - 2017/2/8

Y1 - 2017/2/8

N2 - In learning algorithm design, a direct optimization of a cost function which fits the goal is naturally desired. Particularly in recursive learning, a direct formulation for total-error-rate (TER) minimization is much desired for online classification applications. However, due to a nonlinear counting step in the classification formulation, an exact solution to minimize TER recursively is yet to be established. In this paper, we propose an exact recursive formulation for TER minimization. Based on empirical evaluations using benchmark data sets, we show that the proposed recursive classification algorithm preserves the performance of the batch mode TER while easing the computational memory load by sample based accumulation.

AB - In learning algorithm design, a direct optimization of a cost function which fits the goal is naturally desired. Particularly in recursive learning, a direct formulation for total-error-rate (TER) minimization is much desired for online classification applications. However, due to a nonlinear counting step in the classification formulation, an exact solution to minimize TER recursively is yet to be established. In this paper, we propose an exact recursive formulation for TER minimization. Based on empirical evaluations using benchmark data sets, we show that the proposed recursive classification algorithm preserves the performance of the batch mode TER while easing the computational memory load by sample based accumulation.

UR - http://www.scopus.com/inward/record.url?scp=85015423104&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85015423104&partnerID=8YFLogxK

U2 - 10.1109/TENCON.2016.7847954

DO - 10.1109/TENCON.2016.7847954

M3 - Conference contribution

AN - SCOPUS:85015423104

T3 - IEEE Region 10 Annual International Conference, Proceedings/TENCON

SP - 38

EP - 41

BT - Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2016 IEEE Region 10 Conference, TENCON 2016

Y2 - 22 November 2016 through 25 November 2016

ER -