We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: Fast model-based clustering of partial records

Abstract: Partially recorded data are frequently encountered in many applications and usually clustered by first removing incomplete cases or features with missing values, or by imputing missing values, followed by application of a clustering algorithm to the resulting altered dataset. Here, we develop clustering methodology through a model-based approach using the marginal density for the observed values, assuming a finite mixture model of multivariate $t$ distributions. We compare our approximate algorithm to the corresponding full expectation-maximization (EM) approach that considers the missing values in the incomplete data set and makes a missing at random (MAR) assumption, as well as case deletion and imputation methods. Since only the observed values are utilized, our approach is computationally more efficient than imputation or full EM. Simulation studies demonstrate that our approach has favorable recovery of the true cluster partition compared to case deletion and imputation under various missingness mechanisms, and is at least competitive with the full EM approach, even when MAR assumptions are violated. Our methodology is demonstrated on a problem of clustering gamma-ray bursts and is implemented at this https URL
Comments: 18 pages, 8 figures, 1 table
Subjects: Methodology (stat.ME); High Energy Astrophysical Phenomena (astro-ph.HE); Machine Learning (cs.LG); Computation (stat.CO); Machine Learning (stat.ML)
MSC classes: 62H30, 62G07, 62-04, 62P99
ACM classes: I.5.3
DOI: 10.1002/sta4.416
Cite as: arXiv:2103.16336 [stat.ME]
  (or arXiv:2103.16336v5 [stat.ME] for this version)

Submission history

From: Ranjan Maitra [view email]
[v1] Tue, 30 Mar 2021 13:30:59 GMT (149kb,D)
[v2] Fri, 28 May 2021 21:18:14 GMT (299kb,D)
[v3] Mon, 19 Jul 2021 12:54:29 GMT (2388kb,D)
[v4] Mon, 2 Aug 2021 17:32:15 GMT (4718kb,D)
[v5] Thu, 12 Aug 2021 21:36:33 GMT (4718kb,D)

Link back to: arXiv, form interface, contact.