Uncertainty-Aware Attention for Reliable Interpretation and Prediction

TL;DR

This paper introduces an uncertainty-aware attention mechanism that improves the reliability and interpretability of deep learning models in risk prediction tasks by modeling input-dependent noise and uncertainty.

Contribution

It proposes a novel uncertainty-aware attention mechanism using variational inference, enhancing interpretability and reliability in deep learning models.

Findings

Outperforms existing attention models on electronic health record tasks

Produces attention maps aligned with clinical interpretations

Provides richer interpretation through learned variance and high uncertainty calibration

Abstract

Attention mechanism is effective in both focusing the deep learning models on relevant features and interpreting them. However, attentions may be unreliable since the networks that generate them are often trained in a weakly-supervised manner. To overcome this limitation, we introduce the notion of input-dependent uncertainty to the attention mechanism, such that it generates attention for each feature with varying degrees of noise based on the given input, to learn larger variance on instances it is uncertain about. We learn this Uncertainty-aware Attention (UA) mechanism using variational inference, and validate it on various risk prediction tasks from electronic health records on which our model significantly outperforms existing attention models. The analysis of the learned attentions shows that our model generates attentions that comply with clinicians' interpretation, and provide richer interpretation via learned variance. Further evaluation of both the accuracy of the uncertainty calibration and the prediction performance with "I don't know" decision show that UA yields networks with high reliability as well.