Improved data set characterisation for meta-learning

Y Peng; PA Flach; C Soares; P Brazdil

Improved data set characterisation for meta-learning

Y Peng
, PA Flach
, C Soares
, P Brazdil

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

Abstract

This paper presents new measures, based on the induced decision tree, to characterise datasets for meta-learning in order to select appropriate learning algorithms. The main idea is to capture the characteristics of dataset from the structural shape and size of decision tree induced from the dataset. Totally 15 measures are proposed to describe the structure of a decision tree. Their effectiveness is illustrated through extensive experiments, by comparing to the results obtained by the existing data characteristics techniques, including data characteristics tool (DCT) that is the most wide used technique in meta-learning, and Landmarking that is the most recently developed method.

Translated title of the contribution	Improved data set characterisation for meta-learning
Original language	English
Title of host publication	Unknown
Publisher	Springer
Pages	141 - 152
Number of pages	11
ISBN (Print)	3540001883
Publication status	Published - Jan 2002

Bibliographical note

Conference Proceedings/Title of Journal: Proc. 5th International Conference on Discovery Science (DS-02)

Access to Document

http://www.cs.bris.ac.uk/Publications/pub_info.jsp?id=1000692

Handle.net

Persistent link

Cite this

@inproceedings{1d73905bbc914269b0583c6eb4f3db27,

title = "Improved data set characterisation for meta-learning",

abstract = "This paper presents new measures, based on the induced decision tree, to characterise datasets for meta-learning in order to select appropriate learning algorithms. The main idea is to capture the characteristics of dataset from the structural shape and size of decision tree induced from the dataset. Totally 15 measures are proposed to describe the structure of a decision tree. Their effectiveness is illustrated through extensive experiments, by comparing to the results obtained by the existing data characteristics techniques, including data characteristics tool (DCT) that is the most wide used technique in meta-learning, and Landmarking that is the most recently developed method. ",

author = "Y Peng and PA Flach and C Soares and P Brazdil",

note = "Conference Proceedings/Title of Journal: Proc. 5th International Conference on Discovery Science (DS-02)",

year = "2002",

month = jan,

language = "English",

isbn = "3540001883",

pages = "141 -- 152",

booktitle = "Unknown",

publisher = "Springer",

address = "United States",

}

TY - GEN

T1 - Improved data set characterisation for meta-learning

AU - Peng, Y

AU - Flach, PA

AU - Soares, C

AU - Brazdil, P

N1 - Conference Proceedings/Title of Journal: Proc. 5th International Conference on Discovery Science (DS-02)

PY - 2002/1

Y1 - 2002/1

N2 - This paper presents new measures, based on the induced decision tree, to characterise datasets for meta-learning in order to select appropriate learning algorithms. The main idea is to capture the characteristics of dataset from the structural shape and size of decision tree induced from the dataset. Totally 15 measures are proposed to describe the structure of a decision tree. Their effectiveness is illustrated through extensive experiments, by comparing to the results obtained by the existing data characteristics techniques, including data characteristics tool (DCT) that is the most wide used technique in meta-learning, and Landmarking that is the most recently developed method.

AB - This paper presents new measures, based on the induced decision tree, to characterise datasets for meta-learning in order to select appropriate learning algorithms. The main idea is to capture the characteristics of dataset from the structural shape and size of decision tree induced from the dataset. Totally 15 measures are proposed to describe the structure of a decision tree. Their effectiveness is illustrated through extensive experiments, by comparing to the results obtained by the existing data characteristics techniques, including data characteristics tool (DCT) that is the most wide used technique in meta-learning, and Landmarking that is the most recently developed method.

M3 - Conference Contribution (Conference Proceeding)

SN - 3540001883

SP - 141

EP - 152

BT - Unknown

PB - Springer

ER -

Improved data set characterisation for meta-learning

Abstract

Bibliographical note

Access to Document

Handle.net

Fingerprint

Cite this