Uncertainty-Aware Attention for Reliable Interpretation and Prediction

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.

References in corpus (12)

• Auto-Encoding Variational Bayes
• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• On Calibration of Modern Neural Networks
• The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables
• A Theoretically Grounded Application of Dropout in Recurrent Neural Networks
• End-To-End Memory Networks
• Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference
• Deep and Confident Prediction for Time Series at Uber
• Learning to Detect Sepsis with a Multitask Gaussian Process RNN Classifier
• Benchmark of Deep Learning Models on Large Healthcare MIMIC Datasets
• Bayesian LSTMs in medicine

Cited by in corpus (5)

• A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges
• Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue
• Recent Advances in Fibrosis and Scar Segmentation from Cardiac MRI: A State-of-the-Art Review and Future Perspectives
• Cost-effective Interactive Attention Learning with Neural Attention Processes
• Meta Dropout: Learning to Perturb Features for Generalization