Accelerated classifier training using the PSL cascading structure

Teo Susnjak; Andre L.C. Barczak

doi:10.1007/978-3-642-02490-0_115

Accelerated classifier training using the PSL cascading structure

Teo Susnjak^*, Andre L.C. Barczak

^*Corresponding author for this work

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

Abstract

This paper addresses the problem of excessively long classifier training times associated with using the Adaboost algorithm within the framework of a cascade of boosted ensembles (CoBE). We present new test results confirming the acceleration of the training phase and the robustness of the Parallel Strong classifier within the same Layer (PSL) training structure recently proposed by [1]. The findings demonstrate a speed up of an order of magnitude over the current training methods without a compromise in accuracy. We also present a modified version of the PSL training structure that further decreases the duration of the training phase while preserving accuracy.

Original language	English
Title of host publication	Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers
Editors	Mario Köppen, Nikola Kasabov, George Coghill
Publisher	Springer
Pages	945-952
Number of pages	8
Volume	PART 1
ISBN (Print)	3642024890, 9783642024894
DOIs	https://doi.org/10.1007/978-3-642-02490-0_115
Publication status	Published - 2009
Externally published	Yes
Event	15th International Conference on Neuro-Information Processing, ICONIP 2008 - Auckland, New Zealand Duration: 25 Nov 2008 → 28 Nov 2008 Conference number: 15th

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 1
Volume	5506 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	15th International Conference on Neuro-Information Processing, ICONIP 2008
Abbreviated title	ICONIP 2008
Country/Territory	New Zealand
City	Auckland
Period	25/11/08 → 28/11/08

Access to Document

10.1007/978-3-642-02490-0_115

Cite this

Susnjak, T., & Barczak, A. L. C. (2009). Accelerated classifier training using the PSL cascading structure. In M. Köppen, N. Kasabov, & G. Coghill (Eds.), Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers (Vol. PART 1, pp. 945-952). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5506 LNCS, No. PART 1). Springer. https://doi.org/10.1007/978-3-642-02490-0_115

Susnjak, Teo ; Barczak, Andre L.C. / Accelerated classifier training using the PSL cascading structure. Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers. editor / Mario Köppen ; Nikola Kasabov ; George Coghill. Vol. PART 1 Springer, 2009. pp. 945-952 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).

@inproceedings{5e4ad6d5ca844a16b7687c011b793314,

title = "Accelerated classifier training using the PSL cascading structure",

abstract = "This paper addresses the problem of excessively long classifier training times associated with using the Adaboost algorithm within the framework of a cascade of boosted ensembles (CoBE). We present new test results confirming the acceleration of the training phase and the robustness of the Parallel Strong classifier within the same Layer (PSL) training structure recently proposed by [1]. The findings demonstrate a speed up of an order of magnitude over the current training methods without a compromise in accuracy. We also present a modified version of the PSL training structure that further decreases the duration of the training phase while preserving accuracy.",

author = "Teo Susnjak and Barczak, {Andre L.C.}",

year = "2009",

doi = "10.1007/978-3-642-02490-0_115",

language = "English",

isbn = "3642024890",

volume = "PART 1",

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

publisher = "Springer",

number = "PART 1",

pages = "945--952",

editor = "K{\"o}ppen, {Mario } and Kasabov, {Nikola } and Coghill, {George }",

booktitle = "Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers",

address = "Germany",

note = "15th International Conference on Neuro-Information Processing, ICONIP 2008, ICONIP 2008 ; Conference date: 25-11-2008 Through 28-11-2008",

}

Susnjak, T & Barczak, ALC 2009, Accelerated classifier training using the PSL cascading structure. in M Köppen, N Kasabov & G Coghill (eds), Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers. vol. PART 1, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 5506 LNCS, Springer, pp. 945-952, 15th International Conference on Neuro-Information Processing, ICONIP 2008, Auckland, New Zealand, 25/11/08. https://doi.org/10.1007/978-3-642-02490-0_115

Accelerated classifier training using the PSL cascading structure. / Susnjak, Teo; Barczak, Andre L.C.
Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers. ed. / Mario Köppen; Nikola Kasabov; George Coghill. Vol. PART 1 Springer, 2009. p. 945-952 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5506 LNCS, No. PART 1).

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

TY - GEN

T1 - Accelerated classifier training using the PSL cascading structure

AU - Susnjak, Teo

AU - Barczak, Andre L.C.

N1 - Conference code: 15th

PY - 2009

Y1 - 2009

N2 - This paper addresses the problem of excessively long classifier training times associated with using the Adaboost algorithm within the framework of a cascade of boosted ensembles (CoBE). We present new test results confirming the acceleration of the training phase and the robustness of the Parallel Strong classifier within the same Layer (PSL) training structure recently proposed by [1]. The findings demonstrate a speed up of an order of magnitude over the current training methods without a compromise in accuracy. We also present a modified version of the PSL training structure that further decreases the duration of the training phase while preserving accuracy.

AB - This paper addresses the problem of excessively long classifier training times associated with using the Adaboost algorithm within the framework of a cascade of boosted ensembles (CoBE). We present new test results confirming the acceleration of the training phase and the robustness of the Parallel Strong classifier within the same Layer (PSL) training structure recently proposed by [1]. The findings demonstrate a speed up of an order of magnitude over the current training methods without a compromise in accuracy. We also present a modified version of the PSL training structure that further decreases the duration of the training phase while preserving accuracy.

UR - http://www.scopus.com/inward/record.url?scp=70349128565&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-02490-0_115

DO - 10.1007/978-3-642-02490-0_115

M3 - Conference contribution

AN - SCOPUS:70349128565

SN - 3642024890

SN - 9783642024894

VL - PART 1

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

SP - 945

EP - 952

BT - Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers

A2 - Köppen, Mario

A2 - Kasabov, Nikola

A2 - Coghill, George

PB - Springer

T2 - 15th International Conference on Neuro-Information Processing, ICONIP 2008

Y2 - 25 November 2008 through 28 November 2008

ER -

Susnjak T, Barczak ALC. Accelerated classifier training using the PSL cascading structure. In Köppen M, Kasabov N, Coghill G, editors, Advances in Neuro-Information Processing - 15th International Conference, ICONIP 2008, Revised Selected Papers. Vol. PART 1. Springer. 2009. p. 945-952. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1). doi: 10.1007/978-3-642-02490-0_115

Accelerated classifier training using the PSL cascading structure

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this