PMCRI: a parallel modular classification rule induction framework

F. Stahl; Max Bramer; Mo Adda

doi:10.1007/978-3-642-03070-3_12

PMCRI: a parallel modular classification rule induction framework

F. Stahl, Max Bramer, Mo Adda

School of Computing

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

192 Downloads (Pure)

Abstract

In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.

Original language	English
Title of host publication	Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings
Editors	P. Perner
Place of Publication	Berlin
Publisher	Springer
Pages	148-162
Number of pages	15
Volume	5632
Edition	5632
ISBN (Print)	9783642030697
DOIs	https://doi.org/10.1007/978-3-642-03070-3_12
Publication status	Published - 2009

Publication series

Name	Lecture notes in computer science
Publisher	Springer Veralg
Number	5632
ISSN (Print)	0302-9743

Access to Document

10.1007/978-3-642-03070-3_12

pmcri_A_parallel_modular_classification_rule_induction_framework.pdfSubmitted manuscript, 620 KB

15 Citations
1 Conference contribution

Optimisation of extended generalised fat tree topologies
Peratikou, A. & Adda, M., 2014, Distributed computer and communication networks: 17th international conference, DCCN 2013, Moscow, Russia, October 7-10, 2013. revised selected papers. Vishnevsky, V., Kozyrev, D. & Larionov, A. (eds.). Heidelberg: Springer, p. 82-90 (Communications in computer and information science ; vol. 279).
Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Open Access
File
257 Downloads (Pure)

Cite this

Stahl, F., Bramer, M., & Adda, M. (2009). PMCRI: a parallel modular classification rule induction framework. In P. Perner (Ed.), Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings (5632 ed., Vol. 5632, pp. 148-162). (Lecture notes in computer science; No. 5632). Springer. https://doi.org/10.1007/978-3-642-03070-3_12

Stahl, F. ; Bramer, Max ; Adda, Mo. / PMCRI: a parallel modular classification rule induction framework. Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings. editor / P. Perner. Vol. 5632 5632. ed. Berlin : Springer, 2009. pp. 148-162 (Lecture notes in computer science; 5632).

@inbook{a236178b80544e04a42c91d3eaa330a3,

title = "PMCRI: a parallel modular classification rule induction framework",

abstract = "In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.",

author = "F. Stahl and Max Bramer and Mo Adda",

year = "2009",

doi = "10.1007/978-3-642-03070-3_12",

language = "English",

isbn = "9783642030697",

volume = "5632",

series = "Lecture notes in computer science",

publisher = "Springer",

number = "5632",

pages = "148--162",

editor = "P. Perner",

booktitle = "Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings",

edition = "5632",

}

Stahl, F, Bramer, M & Adda, M 2009, PMCRI: a parallel modular classification rule induction framework. in P Perner (ed.), Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings. 5632 edn, vol. 5632, Lecture notes in computer science, no. 5632, Springer, Berlin, pp. 148-162. https://doi.org/10.1007/978-3-642-03070-3_12

PMCRI: a parallel modular classification rule induction framework. / Stahl, F.; Bramer, Max ; Adda, Mo.
Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings. ed. / P. Perner. Vol. 5632 5632. ed. Berlin: Springer, 2009. p. 148-162 (Lecture notes in computer science; No. 5632).

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

TY - CHAP

T1 - PMCRI: a parallel modular classification rule induction framework

AU - Stahl, F.

AU - Bramer, Max

AU - Adda, Mo

PY - 2009

Y1 - 2009

N2 - In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.

AB - In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.

U2 - 10.1007/978-3-642-03070-3_12

DO - 10.1007/978-3-642-03070-3_12

M3 - Chapter (peer-reviewed)

SN - 9783642030697

VL - 5632

T3 - Lecture notes in computer science

SP - 148

EP - 162

BT - Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings

A2 - Perner, P.

PB - Springer

CY - Berlin

ER -

Stahl F, Bramer M , Adda M. PMCRI: a parallel modular classification rule induction framework. In Perner P, editor, Machine learning and data mining in pattern recognition: 6th international conference, MLDM 2009, Leipzig, Germany, July 23-25, 2009. proceedings. 5632 ed. Vol. 5632. Berlin: Springer. 2009. p. 148-162. (Lecture notes in computer science; 5632). doi: 10.1007/978-3-642-03070-3_12

PMCRI: a parallel modular classification rule induction framework

Abstract

Publication series

Access to Document

Fingerprint

Research output

Optimisation of extended generalised fat tree topologies

Cite this