Cluster analysis of heterogeneous rank data Export

@inproceedings{citeulike:4183688, abstract = {Cluster analysis of ranking data, which occurs in consumer questionnaires, voting forms or other inquiries of preferences, attempts to identify typical groups of rank choices. Empirically measured rankings are often incomplete, i.e. different numbers of filled rank positions cause heterogeneity in the data. We propose a mixture approach for clustering of heterogeneous rank data. Rankings of different lengths can be described and compared by means of a single probabilistic model. A maximum entropy approach avoids hidden assumptions about missing rank positions. Parameter estimators and an efficient EM algorithm for unsupervised inference are derived for the ranking mixture model. Experiments on both synthetic data and real-world data demonstrate significantly improved parameter estimates on heterogeneous data when the incomplete rankings are included in the inference process.}, address = {New York, NY, USA}, author = {Busse, Ludwig M. and Orbanz, Peter and Buhmann, Joachim M.}, booktitle = {ICML '07: Proceedings of the 24th international conference on Machine learning}, citeulike-article-id = {4183688}, citeulike-linkout-0 = {http://portal.acm.org/citation.cfm?id=1273496.1273511}, citeulike-linkout-1 = {http://dx.doi.org/10.1145/1273496.1273511}, doi = {10.1145/1273496.1273511}, isbn = {978-1-59593-793-3}, keywords = {analysis, cluster}, location = {Corvalis, Oregon}, pages = {113--120}, posted-at = {2009-03-16 16:28:33}, priority = {2}, publisher = {ACM}, title = {Cluster analysis of heterogeneous rank data}, url = {http://dx.doi.org/10.1145/1273496.1273511}, year = {2007} }

TY - CONF ID - citeulike:4183688 L3 - citeulike-article-id:4183688 N2 - Cluster analysis of ranking data, which occurs in consumer questionnaires, voting forms or other inquiries of preferences, attempts to identify typical groups of rank choices. Empirically measured rankings are often incomplete, i.e. different numbers of filled rank positions cause heterogeneity in the data. We propose a mixture approach for clustering of heterogeneous rank data. Rankings of different lengths can be described and compared by means of a single probabilistic model. A maximum entropy approach avoids hidden assumptions about missing rank positions. Parameter estimators and an efficient EM algorithm for unsupervised inference are derived for the ranking mixture model. Experiments on both synthetic data and real-world data demonstrate significantly improved parameter estimates on heterogeneous data when the incomplete rankings are included in the inference process. CY - Corvalis, Oregon SN - 978-1-59593-793-3 AD - New York, NY, USA EP - 120 TI - Cluster analysis of heterogeneous rank data PB - ACM T2 - ICML '07: Proceedings of the 24th international conference on Machine learning SP - 113 KW - analysis KW - cluster AU - Busse, Ludwig AU - Orbanz, Peter AU - Buhmann, Joachim PY - 2007/// UR - http://dx.doi.org/10.1145/1273496.1273511 ER -