Abstract
Local learning methods approximate a target function (a posteriori probability) by partitioning the input space into a set of local regions, and modeling a simple input-output relationship in each one. In order for local learning to be effective for pattern classification in high dimensional settings, regions must be chosen judiciously to minimize bias. This paper presents a novel region partitioning criterion that attempts to minimize bias by capturing differential relevance in input variables in an efficient way. The efficacy of the method is validated using a variety of real and simulated data.
Original language | English |
---|---|
Title of host publication | Advances in Pattern Recognition - Joint IAPR International Workshops SSPR 1998 and SPR 1998, Proceedings |
Editors | Adnan Amin, Dov Dori, Pavel Pudil, Herbert Freeman |
Publisher | Springer Verlag |
Pages | 882-889 |
Number of pages | 8 |
ISBN (Print) | 3540648585, 9783540648581 |
State | Published - 1 Jan 1998 |
Event | 7th Joint IAPR International Workshop on Structural and Syntactic Pattern Recognition, SSPR 1998 and 2nd International Workshop on Statistical Techniques in Pattern Recognition, SPR 1998 - Sydney, Australia Duration: 11 Aug 1998 → 13 Aug 1998 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 1451 |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Other
Other | 7th Joint IAPR International Workshop on Structural and Syntactic Pattern Recognition, SSPR 1998 and 2nd International Workshop on Statistical Techniques in Pattern Recognition, SPR 1998 |
---|---|
Country | Australia |
City | Sydney |
Period | 11/08/98 → 13/08/98 |
Fingerprint
Cite this
}
Pattern classification based on local learning. / Peng, Jing; Bhanu, Bir.
Advances in Pattern Recognition - Joint IAPR International Workshops SSPR 1998 and SPR 1998, Proceedings. ed. / Adnan Amin; Dov Dori; Pavel Pudil; Herbert Freeman. Springer Verlag, 1998. p. 882-889 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1451).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Pattern classification based on local learning
AU - Peng, Jing
AU - Bhanu, Bir
PY - 1998/1/1
Y1 - 1998/1/1
N2 - Local learning methods approximate a target function (a posteriori probability) by partitioning the input space into a set of local regions, and modeling a simple input-output relationship in each one. In order for local learning to be effective for pattern classification in high dimensional settings, regions must be chosen judiciously to minimize bias. This paper presents a novel region partitioning criterion that attempts to minimize bias by capturing differential relevance in input variables in an efficient way. The efficacy of the method is validated using a variety of real and simulated data.
AB - Local learning methods approximate a target function (a posteriori probability) by partitioning the input space into a set of local regions, and modeling a simple input-output relationship in each one. In order for local learning to be effective for pattern classification in high dimensional settings, regions must be chosen judiciously to minimize bias. This paper presents a novel region partitioning criterion that attempts to minimize bias by capturing differential relevance in input variables in an efficient way. The efficacy of the method is validated using a variety of real and simulated data.
UR - http://www.scopus.com/inward/record.url?scp=84947777485&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84947777485
SN - 3540648585
SN - 9783540648581
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 882
EP - 889
BT - Advances in Pattern Recognition - Joint IAPR International Workshops SSPR 1998 and SPR 1998, Proceedings
A2 - Amin, Adnan
A2 - Dori, Dov
A2 - Pudil, Pavel
A2 - Freeman, Herbert
PB - Springer Verlag
ER -