About this document
REMASKER: Imputing Tabular Data Efficiently by alejandra.moreno is a document available to read on EtoBox.
The document introduces R E M ASKER, a new method for imputing missing values in tabular data using a masked autoencoding framework. R E M ASKER optimizes an autoencoder by reconstructing both naturally masked and randomly re-masked values, showing strong performance on benchmark datasets compared to existing methods. The findings suggest that masked modeling is a promising direction for future research in tabular data imputation.
- Author
- alejandra.moreno
- Language
- EN