Missing Data: Comparison of Multiple-Imputation Algorithms for Social Determinants of Health in Cervical Cancer Stage Detection

Social Determinants of Health impact general health conditions within a population. However, missing data affect statistical analysis and forecasting of diseases. Multiple imputation has gained momentum and several machine learning algorithms have been used for data imputation. As most statistical a...

Full description

Saved in:
Bibliographic Details
Main Authors: Jaramillo, Juan F. Galindo, Formigari, Giovana Moura, Vale, Diama Bhadra, Ursini, Edson L., Martins, Paulo S.
Format: Conference Proceeding
Language:English
Subjects:
Cervical Cancer
Imputed Data
Machine learning algorithms
MICE
Mobile communication
Multiple Imputation
Predictive models
SDoH
Sociology
Software
Software algorithms
Statistical analysis
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Social Determinants of Health impact general health conditions within a population. However, missing data affect statistical analysis and forecasting of diseases. Multiple imputation has gained momentum and several machine learning algorithms have been used for data imputation. As most statistical analysis and machine learning software have already implemented these algorithms, their performance is usually taken for granted without further analysis. Furthermore, we notice a discrepancy between how imputation must be carried out and how it is usually performed in real-word practice. Thus, in this work we examine different machine learning algorithms for multiple imputation in two datasets with Social Determinants of Health in Cervical Cancer. The results of this comparison are presented.
ISSN:2644-3163
DOI:10.1109/IEMCON53756.2021.9623097