Insightful dimensionality reduction with very low rank variable subsets
Loading...
Access rights
openAccess
Journal Title
Journal ISSN
Volume Title
Conference article in proceedings
This publication is imported from Aalto University research portal.
View publication in the Research portal
View/Open full text file from the Research portal
Other link related to publication
View publication in the Research portal
View/Open full text file from the Research portal
Other link related to publication
Date
2021-06-03
Major/Subject
Mcode
Degree programme
Language
en
Pages
10
3066-3075
3066-3075
Series
Proceedings of the Web Conference, WWW 2021
Abstract
Dimensionality reduction techniques can be employed to produce robust, cost-effective predictive models, and to enhance interpretability in exploratory data analysis. However, the models produced by many of these methods are formulated in terms of abstract factors or are too high-dimensional to facilitate insight and fit within low computational budgets. In this paper we explore an alternative approach to interpretable dimensionality reduction. Given a data matrix, we study the following question: are there subsets of variables that can be primarily explained by a single factor? We formulate this challenge as the problem of finding submatrices close to rank one. Despite its potential, this topic has not been sufficiently addressed in the literature, and there exist virtually no algorithms for this purpose that are simultaneously effective, efficient and scalable. We formalize the task as two problems which we characterize in terms of computational complexity, and propose efficient, scalable algorithms with approximation guarantees. Our experiments demonstrate how our approach can produce insightful findings in data, and show our algorithms to be superior to strong baselines.Description
Funding Information: This work was supported by the Academy of Finland project AIDA (317085), the EC H2020RIA project “SoBigData++” (871042), and the Polish National Agency for Academic Exchange within the Bekker programme, number PPN/BEK/2019/1/00133. Publisher Copyright: © 2021 ACM. | openaire: EC/H2020/871042/EU//SoBigData-PlusPlus
Keywords
Data mining, Dimensionality reduction, Explainability, Variable selection
Other note
Citation
Ordozgoiti , B , Pai , S & Kolczynska , M 2021 , Insightful dimensionality reduction with very low rank variable subsets . in Proceedings of the Web Conference, WWW 2021 . ACM , pp. 3066-3075 , The Web Conference , Ljubljana , Slovenia , 19/04/2021 . https://doi.org/10.1145/3442381.3450067
