CopyMean: a new method to predict monotone missing values in longitudinal studies

CopyMean: a new method to predict monotone missing values in longitudinal studies

Highlights • We present CopyMean, a new method to impute monotone missing values in longitudinal studies. • We compare four version of this new method to 12 classical imputation methods. • For that, we generate some artificial missing values on real data set that had initially no missing data. • Cop...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computer methods and programs in biomedicine 2016-08, Vol.132, p.29-44
Hauptverfasser: Genolini, Christophe, Lacombe, Amandine, Écochard, René, Subtil, Fabien
Format: Artikel
Sprache:eng
Schlagworte:
Alcohol Drinking
Biomarkers
Computer programs
Correlation analysis
Cross-Sectional Studies
Female
Humans
Impact analysis
Imputation
Internal Medicine
Life Sciences
Longitudinal data
Longitudinal Studies
Mathematical models
Menstrual Cycle
Missing data
Other
Statistical analysis
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Highlights • We present CopyMean, a new method to impute monotone missing values in longitudinal studies. • We compare four version of this new method to 12 classical imputation methods. • For that, we generate some artificial missing values on real data set that had initially no missing data. • CopyMean outperforms the other methods in many situations (34 on 54).
ISSN:0169-2607
1872-7565
DOI:10.1016/j.cmpb.2016.04.010