SAINTENS: Self-Attention and Intersample Attention Transformer for Digital Biomarker Development Using Tabular Healthcare Real World Data.
Journal:
Studies in health technology and informatics
Published Date:
May 16, 2022
Abstract
BACKGROUND: Deep learning currently struggles with tabular data, but it can benefit from multimodal learning. SAINT is a deep learning model for tabular data on which we base our presented developments.