Techniques for visualizing LSTMs applied to electrocardiograms
arXiv:1705.08153
Abstract
This paper explores four different visualization techniques for long short-term memory (LSTM) networks applied to continuous-valued time series. On the datasets analysed, we find that the best visualization technique is to learn an input deletion mask that optimally reduces the true class score. With a specific focus on single-lead electrocardiograms from the MIT-BIH arrhythmia dataset, we show that salient input features for the LSTM classifier align well with medical theory.
arXiv commentpresented at 2018 ICML Workshop on Human Interpretability in Machine Learning (WHI 2018), Stockholm, Sweden
References in corpus (13)
- Adam: A Method for Stochastic Optimization
- Sequence to Sequence Learning with Neural Networks
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
- Striving for Simplicity: The All Convolutional Net
- A Critical Review of Recurrent Neural Networks for Sequence Learning
- Visualizing and Understanding Recurrent Networks
- Learning to Diagnose with LSTM Recurrent Neural Networks
- Visualizing and Understanding Neural Models in NLP
- Arrhythmia Classification from the Abductive Interpretation of Short Single-Lead ECG Records
- Architectural Complexity Measures of Recurrent Neural Networks
- Investigating the influence of noise and distractors on the interpretation of neural networks
- Recurrent Batch Normalization
- Genetic Architect: Discovering Genomic Structure with Learned Neural Architectures