Visual Interpretation of Recurrent Neural Network on Multi-dimensional Time-series Forecast

Qiaomu Shen, Yanhong Wu, Yuzhe Jiang, Wei Zeng, Alexis K H LAU, A. Vilanova, Huamin Qu

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

Samenvatting

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 PM 2.5 and SO 2 . 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.
Originele taal-2Engels
Titel2020 IEEE Pacific Visualization Symposium (PacificVis)
RedacteurenFabian Beck, Jinwook Seo, Chaoli Wang
UitgeverijInstitute of Electrical and Electronics Engineers
Pagina's61-70
Aantal pagina's10
ISBN van elektronische versie9781728156972
DOI's
StatusGepubliceerd - jun 2020
Evenement13th IEEE Pacific Visualization Symposium, PacificVis 2020 - Tianjin, China
Duur: 14 apr 202017 apr 2020

Congres

Congres13th IEEE Pacific Visualization Symposium, PacificVis 2020
LandChina
StadTianjin
Periode14/04/2017/04/20

Vingerafdruk Duik in de onderzoeksthema's van 'Visual Interpretation of Recurrent Neural Network on Multi-dimensional Time-series Forecast'. Samen vormen ze een unieke vingerafdruk.

  • Citeer dit

    Shen, Q., Wu, Y., Jiang, Y., Zeng, W., LAU, A. K. H., Vilanova, A., & Qu, H. (2020). Visual Interpretation of Recurrent Neural Network on Multi-dimensional Time-series Forecast. In F. Beck, J. Seo, & C. Wang (editors), 2020 IEEE Pacific Visualization Symposium (PacificVis) (blz. 61-70). [9086238] Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/PacificVis48177.2020.2785