KIT | KIT-Bibliothek | Impressum | Datenschutz

Cluster Correspondence Analysis and Reduced K-Means: A Two-Step Approach to Cluster Low Back Pain Patients

Liu, Fengmei; Gupta, Sucharu; Tortora, Cristina

For the IFCS 2017 data challenge on low back pain (LBP) patients clustering, we used a two-step approach. Two of the challenging characteristics of the data set are the presence of missing values and mixed type variables. After a specific pretreatment, in the first step, we performed domain clustering using cluster correspondence analysis (clusCA). Upon the output variables from each domain, we did the second step, reduced K-means clustering, to get the final clusters of patients. The conclusion section shows the final clustering results and a profile plot of the clusters. Every cluster is highly interpretable and evaluated well with some descriptive variables which are used for measuring the clustering results.

Open Access Logo

Verlagsausgabe §
DOI: 10.5445/KSP/1000085952/03
Veröffentlicht am 14.05.2019
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Informationswirtschaft und Marketing (IISM)
Publikationstyp Zeitschriftenaufsatz
Publikationsjahr 2019
Sprache Englisch
Identifikator ISSN: 2510-0564
KITopen-ID: 1000094555
Erschienen in Archives of Data Science, Series B (Online First)
Band 1
Heft 1
Seiten B03, 29 S. online
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page