Inductive logic programming (ILP) algorithms are
classification algorithms that construct classifiers
represented as logic programs. ILP algorithms have a
number of attractive features, notably the ability to
make use of declarative background (user-supplied)
knowledge. However, ILP algorithms deal poorly with
large data sets (>10000 examples) and their widespread
use of the greedy set-covering algorithm renders them
susceptible to local maxima in the space of logic
programs.
This paper presents a novel approach to address these
problems based on combining the local search properties
of an inductive logic programming algorithm with the
global search properties of an evolutionary algorithm.
The proposed algorithm may be viewed as an evolutionary
wrapper around a population of ILP algorithms.
The evolutionary wrapper approach is evaluated on two
domains. The chess-endgame (KRK) problem is an
artificial domain that is a widely used benchmark in
inductive logic programming, and Part-of-Speech Tagging
is a real-world problem from the field of Natural
Language Processing. In the latter domain, data
originates from excerpts of the Wall Street Journal.
Results indicate that significant improvements in
predictive accuracy can be achieved over a conventional
ILP approach when data is plentiful and noisy.
%0 Journal Article
%1 Reiser:2001:AI
%A Reiser, Philip G. K.
%A Riddle, Patricia J.
%D 2001
%J Applied Intelligence
%K ILP algorithms, evolutionary genetic inductive learning, logic machine programming, sampling,
%N 3
%P 181--197
%R doi:10.1023/A:1011239013893
%T Scaling Up Inductive Logic Programming: An
Evolutionary Wrapper Approach
%U http://www.stancomb.co.uk/~prr/Papers/AppInt.ps
%V 15
%X Inductive logic programming (ILP) algorithms are
classification algorithms that construct classifiers
represented as logic programs. ILP algorithms have a
number of attractive features, notably the ability to
make use of declarative background (user-supplied)
knowledge. However, ILP algorithms deal poorly with
large data sets (>10000 examples) and their widespread
use of the greedy set-covering algorithm renders them
susceptible to local maxima in the space of logic
programs.
This paper presents a novel approach to address these
problems based on combining the local search properties
of an inductive logic programming algorithm with the
global search properties of an evolutionary algorithm.
The proposed algorithm may be viewed as an evolutionary
wrapper around a population of ILP algorithms.
The evolutionary wrapper approach is evaluated on two
domains. The chess-endgame (KRK) problem is an
artificial domain that is a widely used benchmark in
inductive logic programming, and Part-of-Speech Tagging
is a real-world problem from the field of Natural
Language Processing. In the latter domain, data
originates from excerpts of the Wall Street Journal.
Results indicate that significant improvements in
predictive accuracy can be achieved over a conventional
ILP approach when data is plentiful and noisy.
@article{Reiser:2001:AI,
abstract = {Inductive logic programming (ILP) algorithms are
classification algorithms that construct classifiers
represented as logic programs. ILP algorithms have a
number of attractive features, notably the ability to
make use of declarative background (user-supplied)
knowledge. However, ILP algorithms deal poorly with
large data sets (>10000 examples) and their widespread
use of the greedy set-covering algorithm renders them
susceptible to local maxima in the space of logic
programs.
This paper presents a novel approach to address these
problems based on combining the local search properties
of an inductive logic programming algorithm with the
global search properties of an evolutionary algorithm.
The proposed algorithm may be viewed as an evolutionary
wrapper around a population of ILP algorithms.
The evolutionary wrapper approach is evaluated on two
domains. The chess-endgame (KRK) problem is an
artificial domain that is a widely used benchmark in
inductive logic programming, and Part-of-Speech Tagging
is a real-world problem from the field of Natural
Language Processing. In the latter domain, data
originates from excerpts of the Wall Street Journal.
Results indicate that significant improvements in
predictive accuracy can be achieved over a conventional
ILP approach when data is plentiful and noisy.},
added-at = {2008-06-19T17:46:40.000+0200},
author = {Reiser, Philip G. K. and Riddle, Patricia J.},
biburl = {https://www.bibsonomy.org/bibtex/2dc6d1615bf289b160eb06fc939a6e3f3/brazovayeye},
doi = {doi:10.1023/A:1011239013893},
interhash = {013afb227c7434aa3898e4d128732ab5},
intrahash = {dc6d1615bf289b160eb06fc939a6e3f3},
issn = {0924-669X},
journal = {Applied Intelligence},
keywords = {ILP algorithms, evolutionary genetic inductive learning, logic machine programming, sampling,},
month = {November-December},
note = {Special Issue: Simulated Evolution and Learning},
notes = {Article ID: 354285
Crossover},
number = 3,
pages = {181--197},
size = {17 pages},
timestamp = {2008-06-19T17:50:11.000+0200},
title = {Scaling Up Inductive Logic Programming: An
Evolutionary Wrapper Approach},
url = {http://www.stancomb.co.uk/~prr/Papers/AppInt.ps},
volume = 15,
year = 2001
}