NeuroView-RNN: It's About Time
Not on the map. Why?
Abstract (arXiv)
Recurrent Neural Networks (RNNs) are important tools for processing sequential data such as time-series or video. Interpretability is defined as the ability to be understood by a person and is different from explainability, which is the ability to be explained in a mathematical formulation. A key interpretability issue with RNNs is that it is not clear how each hidden state per time step contributes to the decision-making process in a quantitative manner. We propose NeuroView-RNN as a family of new RNN architectures that explains how all the time steps are used for the decision-making process. Each member of the family is derived from a standard RNN architecture by concatenation of the hidden steps into a global linear classifier. The global linear classifier has all the hidden states as the input, so the weights of the classifier have a linear mapping to the hidden states. Hence, from the weights, NeuroView-RNN can quantify how important each time step is to a particular decision. As a bonus, NeuroView-RNN also offers higher accuracy in many cases compared to the RNNs and their variants. We showcase the benefits of NeuroView-RNN by evaluating on a multitude of diverse time-series datasets.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 28 of its 48 references that are in the map. First posted on arXiv 23 Feb 2022.
Builds on
28 of 48 references.
References · 28
- 2013Efficient Estimation of Word Representations in Vector Spacecited by 4,346
- 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,016
- 2014Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modelingcited by 2,341
- 2016Enriching Word Vectors with Subword Informationcited by 1,578
- 2018Explaining Explanations: An Overview of Interpretability of Machine Learningcited by 508
- 2020Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representationscited by 154
- 2020Learning from Failure: Training Debiased Classifier from Biased Classifiercited by 133
- 2015A Simple Way to Initialize Recurrent Networks of Rectified Linear Unitscited by 130
- 2017Dipole: Diagnosis Prediction in Healthcare via Attention-based Bidirectional Recurrent Neural Networkscited by 114
- 2019Interpretable and Steerable Sequence Learning via Prototypescited by 72
- 2020The Recurrent Neural Tangent Kernelcited by 67
- 2013How to Construct Deep Recurrent Neural Networkscited by 64
16 more
- 2019Exploring Interpretable LSTM Neural Networks over Multi-Variable Datacited by 42
- 2021Human Trajectory Prediction via Counterfactual Analysiscited by 42
- 2018A Comparison of Five Multiple Instance Learning Pooling Functions for Sound Event Detection with Weak Labelingcited by 37
- 2017SMILES2Vec: An Interpretable General-Purpose Deep Neural Network for Predicting Chemical Propertiescited by 30
- 2019Explaining and Interpreting LSTMscited by 23
- 2016Image Captioning with Deep Bidirectional LSTMscited by 21
- 2018Interpretable Spatio-temporal Attention for Video Action Recognitioncited by 20
- 2017Two Stream LSTM: A Deep Fusion Framework for Human Action Recognitioncited by 18
- 2017From Deterministic to Generative: Multi-Modal Stochastic RNNs for Video Captioningcited by 15
- 2017Improving Interpretability of Deep Neural Networks with Semantic Informationcited by 14
- 2015Improving performance of recurrent neural network with relu nonlinearitycited by 12
- 2017Beat by Beat: Classifying Cardiac Arrhythmias with Recurrent Neural Networkscited by 5
- 2017RNN-based counterfactual prediction, with an application to homestead policy and public schoolingcited by 4
- 2020A Comparison of Pooling Methods on LSTM Models for Rare Acoustic Event Classificationcited by 3
- 2016Empirical Evaluation of RNN Architectures on Sentence Classification Taskcited by 1
- 2021NeuroView: Explainable Deep Network Decision Makingcited by 1
Built on by
Citing papers per year: 2023: 1; 2024: 0; 2025: 1.
2 papers.