Print Email Facebook Twitter Visual Interpretation of Recurrent Neural Network on Multi-dimensional Time-series Forecast Title Visual Interpretation of Recurrent Neural Network on Multi-dimensional Time-series Forecast Author Shen, Qiaomu (The Hong Kong University of Science and Technology) Wu, Yanhong (Visa Research) Jiang, Yuzhe (The Hong Kong University of Science and Technology) Zeng, Wei (Shenzhen Institute of Advanced Technologies) Lau, Alexis K.H. (The Hong Kong University of Science and Technology) Vilanova Bartroli, A. (TU Delft Computer Graphics and Visualisation) Qu, Huamin (The Hong Kong University of Science and Technology) Contributor Beck, Fabian (editor) Seo, Jinwook (editor) Wang, Chaoli (editor) Date 2020 Abstract Recent attempts at utilizing visual analytics to interpret Recurrent Neural Networks (RNNs) mainly focus on natural language processing (NLP) tasks that take symbolic sequences as input. However, many real-world problems like environment pollution forecasting apply RNNs on sequences of multi-dimensional data where each dimension represents an individual feature with semantic meaning such as PM2.5 and SO2. RNN interpretation on multi-dimensional sequences is challenging as users need to analyze what features are important at different time steps to better understand model behavior and gain trust in prediction. This requires effective and scalable visualization methods to reveal the complex many-to-many relations between hidden units and features. In this work, we propose a visual analytics system to interpret RNNs on multi-dimensional time-series forecasts. Specifically, to provide an overview to reveal the model mechanism, we propose a technique to estimate the hidden unit response by measuring how different feature selections affect the hidden unit output distribution. We then cluster the hidden units and features based on the response embedding vectors. Finally, we propose a visual analytics system which allows users to visually explore the model behavior from the global and individual levels. We demonstrate the effectiveness of our approach with case studies using air pollutant forecast applications. Subject air pollutant forecastinterpretable machine learningmulti-dimensional time seriesrecurrent neural networks To reference this document use: http://resolver.tudelft.nl/uuid:ed13ced1-14e4-4eea-84ed-a677da2bd900 DOI https://doi.org/10.1109/PacificVis48177.2020.2785 Publisher IEEE ISBN 9781728156972 Source 2020 IEEE Pacific Visualization Symposium, PacificVis 2020 - Proceedings, 2020-June Event 13th IEEE Pacific Visualization Symposium, PacificVis 2020, 2020-04-14 → 2020-04-17, Tianjin, China Bibliographical note Accepted author manuscript Part of collection Institutional Repository Document type conference paper Rights © 2020 Qiaomu Shen, Yanhong Wu, Yuzhe Jiang, Wei Zeng, Alexis K.H. Lau, A. Vilanova Bartroli, Huamin Qu Files PDF pacificvis20a_sub2785_cam_i5.pdf 7.18 MB Close viewer /islandora/object/uuid:ed13ced1-14e4-4eea-84ed-a677da2bd900/datastream/OBJ/view