Discriminative clustering for market segmentation

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Standard

Discriminative clustering for market segmentation. / Haider, Peter; Chiarandini, Luca; Brefeld, Ulf.
Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining. New York: Association for Computing Machinery, Inc, 2012. p. 417-425.

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Harvard

Haider, P, Chiarandini, L & Brefeld, U 2012, Discriminative clustering for market segmentation. in Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining. Association for Computing Machinery, Inc, New York, pp. 417-425, 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD-2012, Beijing, China, 12.08.12. https://doi.org/10.1145/2339530.2339600

APA

Haider, P., Chiarandini, L., & Brefeld, U. (2012). Discriminative clustering for market segmentation. In Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 417-425). Association for Computing Machinery, Inc. https://doi.org/10.1145/2339530.2339600

Vancouver

Haider P, Chiarandini L, Brefeld U. Discriminative clustering for market segmentation. In Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining. New York: Association for Computing Machinery, Inc. 2012. p. 417-425 doi: 10.1145/2339530.2339600

Bibtex

@inbook{5e0e09eff4a641fcb14544bc4ddc0bfd,
title = "Discriminative clustering for market segmentation",
abstract = "We study discriminative clustering for market segmentation tasks. The underlying problem setting resembles discriminative clustering, however, existing approaches focus on the prediction of univariate cluster labels. By contrast, market segments encode complex (future) behavior of the individuals which cannot be represented by a single variable. In this paper, we generalize discriminative clustering to structured and complex output variables that can be represented as graphical models. We devise two novel methods to jointly learn the classifier and the clustering using alternating optimization and collapsed inference, respectively. The two approaches jointly learn a discriminative segmentation of the input space and a generative output prediction model for each segment. We evaluate our methods on segmenting user navigation sequences from Yahoo! News. The proposed collapsed algorithm is observed to outperform baseline approaches such as mixture of experts. We showcase exemplary projections of the resulting segments to display the interpretability of the solutions. ",
keywords = "Informatics, Alternating optimizations, Discriminative clustering , GraphicaL model, Input space, Interpretability, Market segment, Market segmentation, Mixture of experts, Novel methods, Output variables, Prediction model, Single variable, Univariate, User navigation, Business informatics",
author = "Peter Haider and Luca Chiarandini and Ulf Brefeld",
year = "2012",
doi = "10.1145/2339530.2339600",
language = "English",
isbn = "978-1-4503-1462-6",
pages = "417--425",
booktitle = "Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining",
publisher = "Association for Computing Machinery, Inc",
address = "United States",
note = "18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD-2012 : Mining the Big Data, KDD 2012 ; Conference date: 12-08-2012 Through 16-08-2012",
url = "http://kdd2012.sigkdd.org/",

}

RIS

TY - CHAP

T1 - Discriminative clustering for market segmentation

AU - Haider, Peter

AU - Chiarandini, Luca

AU - Brefeld, Ulf

N1 - Conference code: 18

PY - 2012

Y1 - 2012

N2 - We study discriminative clustering for market segmentation tasks. The underlying problem setting resembles discriminative clustering, however, existing approaches focus on the prediction of univariate cluster labels. By contrast, market segments encode complex (future) behavior of the individuals which cannot be represented by a single variable. In this paper, we generalize discriminative clustering to structured and complex output variables that can be represented as graphical models. We devise two novel methods to jointly learn the classifier and the clustering using alternating optimization and collapsed inference, respectively. The two approaches jointly learn a discriminative segmentation of the input space and a generative output prediction model for each segment. We evaluate our methods on segmenting user navigation sequences from Yahoo! News. The proposed collapsed algorithm is observed to outperform baseline approaches such as mixture of experts. We showcase exemplary projections of the resulting segments to display the interpretability of the solutions.

AB - We study discriminative clustering for market segmentation tasks. The underlying problem setting resembles discriminative clustering, however, existing approaches focus on the prediction of univariate cluster labels. By contrast, market segments encode complex (future) behavior of the individuals which cannot be represented by a single variable. In this paper, we generalize discriminative clustering to structured and complex output variables that can be represented as graphical models. We devise two novel methods to jointly learn the classifier and the clustering using alternating optimization and collapsed inference, respectively. The two approaches jointly learn a discriminative segmentation of the input space and a generative output prediction model for each segment. We evaluate our methods on segmenting user navigation sequences from Yahoo! News. The proposed collapsed algorithm is observed to outperform baseline approaches such as mixture of experts. We showcase exemplary projections of the resulting segments to display the interpretability of the solutions.

KW - Informatics

KW - Alternating optimizations

KW - Discriminative clustering

KW - GraphicaL model

KW - Input space

KW - Interpretability

KW - Market segment

KW - Market segmentation

KW - Mixture of experts

KW - Novel methods

KW - Output variables

KW - Prediction model

KW - Single variable

KW - Univariate

KW - User navigation

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=84866039328&partnerID=8YFLogxK

U2 - 10.1145/2339530.2339600

DO - 10.1145/2339530.2339600

M3 - Article in conference proceedings

AN - SCOPUS:84866039328

SN - 978-1-4503-1462-6

SP - 417

EP - 425

BT - Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining

PB - Association for Computing Machinery, Inc

CY - New York

T2 - 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD-2012

Y2 - 12 August 2012 through 16 August 2012

ER -

DOI