Artikel

High dimensional, robust, unsupervised record linkage

We develop a technique for record linkage on high dimensional data, where the two datasets may not have any common variable, and there may be no training set available. Our methodology is based on sparse, high dimensional principal components. Since large and high dimensional datasets are often prone to outliers and aberrant observations, we propose a technique for estimating robust, high dimensional principal components. We present theoretical results validating the robust, high dimensional principal component estimation steps, and justifying their use for record linkage. Some numeric results and remarks are also presented.

Language
Englisch

Bibliographic citation
Journal: Statistics in Transition New Series ; ISSN: 2450-0291 ; Volume: 21 ; Year: 2020 ; Issue: 4 ; Pages: 123-143 ; New York: Exeley

Subject
record linkage
principal components
high dimensional
robust

Event
Geistige Schöpfung
(who)
Bera, Sabyasachi
Chatterjee, Snigdhansu
Event
Veröffentlichung
(who)
Exeley
(where)
New York
(when)
2020

DOI
doi:10.21307/stattrans-2020-034
Handle
Last update
10.03.2025, 11:43 AM CET

Data provider

This object is provided by:
ZBW - Deutsche Zentralbibliothek für Wirtschaftswissenschaften - Leibniz-Informationszentrum Wirtschaft. If you have any questions about the object, please contact the data provider.

Object type

  • Artikel

Associated

  • Bera, Sabyasachi
  • Chatterjee, Snigdhansu
  • Exeley

Time of origin

  • 2020

Other Objects (12)