Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

SCOAL: A framework for simultaneous co-clustering and learning from complex data

SCOAL: A framework for simultaneous co-clustering and learning from complex data SCOAL: A Framework for Simultaneous Co-Clustering and Learning from Complex Data MEGHANA DEODHAR and JOYDEEP GHOSH University of Texas at Austin For dif cult classi cation or regression problems, practitioners often segment the data into relatively homogeneous groups and then build a predictive model for each group. This two-step procedure usually results in simpler, more interpretable and actionable models without any loss in accuracy. In this work, we consider problems such as predicting customer behavior across products, where the independent variables can be naturally partitioned into two sets, that is, the data is dyadic in nature. A pivoting operation now results in the dependent variable showing up as entries in a œcustomer by product  data matrix. We present the Simultaneous CO-clustering And Learning (SCOAL) framework, based on the key idea of interleaving co-clustering and construction of prediction models to iteratively improve both cluster assignment and t of the models. This algorithm provably converges to a local minimum of a suitable cost function. The framework not only generalizes co-clustering and collaborative ltering to model-based co-clustering, but can also be viewed as simultaneous co-segmentation and classi cation or regression, which is typically better than independently clustering the data rst http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png ACM Transactions on Knowledge Discovery from Data (TKDD) Association for Computing Machinery

SCOAL: A framework for simultaneous co-clustering and learning from complex data

Loading next page...
 
/lp/association-for-computing-machinery/scoal-a-framework-for-simultaneous-co-clustering-and-learning-from-Jn0ImJOyzz

References (36)

Publisher
Association for Computing Machinery
Copyright
Copyright © 2010 by ACM Inc.
ISSN
1556-4681
DOI
10.1145/1839490.1839492
Publisher site
See Article on Publisher Site

Abstract

SCOAL: A Framework for Simultaneous Co-Clustering and Learning from Complex Data MEGHANA DEODHAR and JOYDEEP GHOSH University of Texas at Austin For dif cult classi cation or regression problems, practitioners often segment the data into relatively homogeneous groups and then build a predictive model for each group. This two-step procedure usually results in simpler, more interpretable and actionable models without any loss in accuracy. In this work, we consider problems such as predicting customer behavior across products, where the independent variables can be naturally partitioned into two sets, that is, the data is dyadic in nature. A pivoting operation now results in the dependent variable showing up as entries in a œcustomer by product  data matrix. We present the Simultaneous CO-clustering And Learning (SCOAL) framework, based on the key idea of interleaving co-clustering and construction of prediction models to iteratively improve both cluster assignment and t of the models. This algorithm provably converges to a local minimum of a suitable cost function. The framework not only generalizes co-clustering and collaborative ltering to model-based co-clustering, but can also be viewed as simultaneous co-segmentation and classi cation or regression, which is typically better than independently clustering the data rst

Journal

ACM Transactions on Knowledge Discovery from Data (TKDD)Association for Computing Machinery

Published: Oct 1, 2010

There are no references for this article.