The role of feature construction in inductive rule learning

Peter Flach; Raedt Luc De; Lavrac Nada; Kramer Stefan

The role of feature construction in inductive rule learning

Peter Flach
, Raedt Luc De
, Lavrac Nada
, Kramer Stefan

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

Abstract

This paper proposes a unifying framework for inductive rule learning algorithms. We suggest that the problem of constructing an appropriate inductive hypothesis (set of rules) can be broken down in the following subtasks: rule construction, body construction, and feature construction. Each of these subtasks may have its own declarative bias, search strategies, and heuristics. In particular, we argue that feature construction is a crucial notion in explaining the relations between attribute-value rule learning and inductive logic programming (ILP). We demonstrate this by a general method for transforming ILP problems to attribute-value form, which overcomes some of the traditional limitations of propositionalisation approaches.

Translated title of the contribution	The role of feature construction in inductive rule learning
Original language	English
Title of host publication	Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries
Publisher	17th International Conference on Machine Learning
Pages	1 - 11
Number of pages	10
Publication status	Published - 2000

Bibliographical note

Other page information: 1-11
Conference Proceedings/Title of Journal: Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries
Other identifier: 1000486

Access to Document

http://www.cs.bris.ac.uk/Publications/pub_master.jsp?id=1000486

Handle.net

Persistent link

Cite this

@inproceedings{1141dff9f8944dcf887cc0f14b585324,

title = "The role of feature construction in inductive rule learning",

abstract = "This paper proposes a unifying framework for inductive rule learning algorithms. We suggest that the problem of constructing an appropriate inductive hypothesis (set of rules) can be broken down in the following subtasks: rule construction, body construction, and feature construction. Each of these subtasks may have its own declarative bias, search strategies, and heuristics. In particular, we argue that feature construction is a crucial notion in explaining the relations between attribute-value rule learning and inductive logic programming (ILP). We demonstrate this by a general method for transforming ILP problems to attribute-value form, which overcomes some of the traditional limitations of propositionalisation approaches. ",

author = "Peter Flach and \{Luc De\}, Raedt and Lavrac Nada and Kramer Stefan",

note = "Other page information: 1-11 Conference Proceedings/Title of Journal: Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries Other identifier: 1000486",

year = "2000",

language = "English",

pages = "1 -- 11",

booktitle = "Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries",

publisher = "17th International Conference on Machine Learning",

}

The role of feature construction in inductive rule learning. / Flach, Peter; Luc De, Raedt; Nada, Lavrac et al.
Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries. 17th International Conference on Machine Learning, 2000. p. 1 - 11.

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

TY - GEN

T1 - The role of feature construction in inductive rule learning

AU - Flach, Peter

AU - Luc De, Raedt

AU - Nada, Lavrac

AU - Stefan, Kramer

N1 - Other page information: 1-11 Conference Proceedings/Title of Journal: Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries Other identifier: 1000486

PY - 2000

Y1 - 2000

N2 - This paper proposes a unifying framework for inductive rule learning algorithms. We suggest that the problem of constructing an appropriate inductive hypothesis (set of rules) can be broken down in the following subtasks: rule construction, body construction, and feature construction. Each of these subtasks may have its own declarative bias, search strategies, and heuristics. In particular, we argue that feature construction is a crucial notion in explaining the relations between attribute-value rule learning and inductive logic programming (ILP). We demonstrate this by a general method for transforming ILP problems to attribute-value form, which overcomes some of the traditional limitations of propositionalisation approaches.

AB - This paper proposes a unifying framework for inductive rule learning algorithms. We suggest that the problem of constructing an appropriate inductive hypothesis (set of rules) can be broken down in the following subtasks: rule construction, body construction, and feature construction. Each of these subtasks may have its own declarative bias, search strategies, and heuristics. In particular, we argue that feature construction is a crucial notion in explaining the relations between attribute-value rule learning and inductive logic programming (ILP). We demonstrate this by a general method for transforming ILP problems to attribute-value form, which overcomes some of the traditional limitations of propositionalisation approaches.

M3 - Conference Contribution (Conference Proceeding)

SP - 1

EP - 11

BT - Proceedings of the ICML2000 workshop on Attribute-Value and Relational Learning: crossing the boundaries

PB - 17th International Conference on Machine Learning

ER -

The role of feature construction in inductive rule learning

Abstract

Bibliographical note

Access to Document

Handle.net

Fingerprint

Cite this