1 Citation (Scopus)

Abstract

Two linked ensembles are used for a supervised learning problem with rare-event counts. With many target instances of zero, more traditional loss functions (such as squared error and class error) are often not relevant and a statistical model leads to a likelihood with two related parameters from a zero-inflated Poisson (ZIP) distribution. In a new approach, a linked pair of gradient boosted tree ensembles are developed to handle the multiple parameters in a manner that can be generalized to other problems. The result is a unique learner that extends machine learning methods to data with nontraditional structures. We empirically compare to two real data sets and two artificial data sets versus a single-tree approach (ZIP-tree) and a statistical generalized linear model.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages225-236
Number of pages12
Volume5772 LCNS
DOIs
StatePublished - 2009
Event8th International Symposium on Intelligent Data Analysis, IDA 2009 - Lyon, France
Duration: Aug 31 2009Sep 2 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5772 LCNS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other8th International Symposium on Intelligent Data Analysis, IDA 2009
CountryFrance
CityLyon
Period8/31/099/2/09

Fingerprint

Rare Events
Count
Ensemble
Poisson distribution
Supervised learning
Zero
Zero Distribution
Learning systems
Generalized Linear Model
Supervised Learning
Loss Function
Statistical Model
Likelihood
Siméon Denis Poisson
Machine Learning
Gradient
Target
Statistical Models

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Borisov, A., Runger, G., Tuv, E., & Lurponglukana-Strand, N. (2009). Zero-inflated boosted ensembles for rare event counts. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5772 LCNS, pp. 225-236). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5772 LCNS). https://doi.org/10.1007/978-3-642-03915-7_20

Zero-inflated boosted ensembles for rare event counts. / Borisov, Alexander; Runger, George; Tuv, Eugene; Lurponglukana-Strand, Nuttha.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5772 LCNS 2009. p. 225-236 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5772 LCNS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Borisov, A, Runger, G, Tuv, E & Lurponglukana-Strand, N 2009, Zero-inflated boosted ensembles for rare event counts. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 5772 LCNS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5772 LCNS, pp. 225-236, 8th International Symposium on Intelligent Data Analysis, IDA 2009, Lyon, France, 8/31/09. https://doi.org/10.1007/978-3-642-03915-7_20
Borisov A, Runger G, Tuv E, Lurponglukana-Strand N. Zero-inflated boosted ensembles for rare event counts. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5772 LCNS. 2009. p. 225-236. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-03915-7_20
Borisov, Alexander ; Runger, George ; Tuv, Eugene ; Lurponglukana-Strand, Nuttha. / Zero-inflated boosted ensembles for rare event counts. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5772 LCNS 2009. pp. 225-236 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{c1ac1aaabef54d479166f783bcefe309,
title = "Zero-inflated boosted ensembles for rare event counts",
abstract = "Two linked ensembles are used for a supervised learning problem with rare-event counts. With many target instances of zero, more traditional loss functions (such as squared error and class error) are often not relevant and a statistical model leads to a likelihood with two related parameters from a zero-inflated Poisson (ZIP) distribution. In a new approach, a linked pair of gradient boosted tree ensembles are developed to handle the multiple parameters in a manner that can be generalized to other problems. The result is a unique learner that extends machine learning methods to data with nontraditional structures. We empirically compare to two real data sets and two artificial data sets versus a single-tree approach (ZIP-tree) and a statistical generalized linear model.",
author = "Alexander Borisov and George Runger and Eugene Tuv and Nuttha Lurponglukana-Strand",
year = "2009",
doi = "10.1007/978-3-642-03915-7_20",
language = "English (US)",
isbn = "3642039146",
volume = "5772 LCNS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "225--236",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Zero-inflated boosted ensembles for rare event counts

AU - Borisov, Alexander

AU - Runger, George

AU - Tuv, Eugene

AU - Lurponglukana-Strand, Nuttha

PY - 2009

Y1 - 2009

N2 - Two linked ensembles are used for a supervised learning problem with rare-event counts. With many target instances of zero, more traditional loss functions (such as squared error and class error) are often not relevant and a statistical model leads to a likelihood with two related parameters from a zero-inflated Poisson (ZIP) distribution. In a new approach, a linked pair of gradient boosted tree ensembles are developed to handle the multiple parameters in a manner that can be generalized to other problems. The result is a unique learner that extends machine learning methods to data with nontraditional structures. We empirically compare to two real data sets and two artificial data sets versus a single-tree approach (ZIP-tree) and a statistical generalized linear model.

AB - Two linked ensembles are used for a supervised learning problem with rare-event counts. With many target instances of zero, more traditional loss functions (such as squared error and class error) are often not relevant and a statistical model leads to a likelihood with two related parameters from a zero-inflated Poisson (ZIP) distribution. In a new approach, a linked pair of gradient boosted tree ensembles are developed to handle the multiple parameters in a manner that can be generalized to other problems. The result is a unique learner that extends machine learning methods to data with nontraditional structures. We empirically compare to two real data sets and two artificial data sets versus a single-tree approach (ZIP-tree) and a statistical generalized linear model.

UR - http://www.scopus.com/inward/record.url?scp=70349855115&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349855115&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-03915-7_20

DO - 10.1007/978-3-642-03915-7_20

M3 - Conference contribution

AN - SCOPUS:70349855115

SN - 3642039146

SN - 9783642039140

VL - 5772 LCNS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 225

EP - 236

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -