Recurrent Neural Networks for Multivariate Time Series with Missing Values
© 2018 The Author(s). Multivariate time series data in practical applications, such as health care, geoscience, and biology, are characterized by a variety of missing values. In time series prediction and other related tasks, it has been noted that missing values and their missing patterns are often...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer Nature
2021
|
| Online Access: | https://hdl.handle.net/1721.1/134960 |
| _version_ | 1811088831845236736 |
|---|---|
| author | Che, Zhengping Purushotham, Sanjay Cho, Kyunghyun Sontag, David Liu, Yan |
| author2 | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science |
| author_facet | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Che, Zhengping Purushotham, Sanjay Cho, Kyunghyun Sontag, David Liu, Yan |
| author_sort | Che, Zhengping |
| collection | MIT |
| description | © 2018 The Author(s). Multivariate time series data in practical applications, such as health care, geoscience, and biology, are characterized by a variety of missing values. In time series prediction and other related tasks, it has been noted that missing values and their missing patterns are often correlated with the target labels, a.k.a., informative missingness. There is very limited work on exploiting the missing patterns for effective imputation and improving prediction performance. In this paper, we develop novel deep learning models, namely GRU-D, as one of the early attempts. GRU-D is based on Gated Recurrent Unit (GRU), a state-of-the-art recurrent neural network. It takes two representations of missing patterns, i.e., masking and time interval, and effectively incorporates them into a deep model architecture so that it not only captures the long-term temporal dependencies in time series, but also utilizes the missing patterns to achieve better prediction results. Experiments of time series classification tasks on real-world clinical datasets (MIMIC-III, PhysioNet) and synthetic datasets demonstrate that our models achieve state-of-the-art performance and provide useful insights for better understanding and utilization of missing values in time series analysis. |
| first_indexed | 2024-09-23T14:08:11Z |
| format | Article |
| id | mit-1721.1/134960 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T14:08:11Z |
| publishDate | 2021 |
| publisher | Springer Nature |
| record_format | dspace |
| spelling | mit-1721.1/1349602023-02-23T20:40:30Z Recurrent Neural Networks for Multivariate Time Series with Missing Values Che, Zhengping Purushotham, Sanjay Cho, Kyunghyun Sontag, David Liu, Yan Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science © 2018 The Author(s). Multivariate time series data in practical applications, such as health care, geoscience, and biology, are characterized by a variety of missing values. In time series prediction and other related tasks, it has been noted that missing values and their missing patterns are often correlated with the target labels, a.k.a., informative missingness. There is very limited work on exploiting the missing patterns for effective imputation and improving prediction performance. In this paper, we develop novel deep learning models, namely GRU-D, as one of the early attempts. GRU-D is based on Gated Recurrent Unit (GRU), a state-of-the-art recurrent neural network. It takes two representations of missing patterns, i.e., masking and time interval, and effectively incorporates them into a deep model architecture so that it not only captures the long-term temporal dependencies in time series, but also utilizes the missing patterns to achieve better prediction results. Experiments of time series classification tasks on real-world clinical datasets (MIMIC-III, PhysioNet) and synthetic datasets demonstrate that our models achieve state-of-the-art performance and provide useful insights for better understanding and utilization of missing values in time series analysis. 2021-10-27T20:10:03Z 2021-10-27T20:10:03Z 2018 2019-07-03T15:21:47Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/134960 en 10.1038/S41598-018-24271-9 Scientific Reports Creative Commons Attribution 4.0 International license https://creativecommons.org/licenses/by/4.0/ application/pdf Springer Nature Nature |
| spellingShingle | Che, Zhengping Purushotham, Sanjay Cho, Kyunghyun Sontag, David Liu, Yan Recurrent Neural Networks for Multivariate Time Series with Missing Values |
| title | Recurrent Neural Networks for Multivariate Time Series with Missing Values |
| title_full | Recurrent Neural Networks for Multivariate Time Series with Missing Values |
| title_fullStr | Recurrent Neural Networks for Multivariate Time Series with Missing Values |
| title_full_unstemmed | Recurrent Neural Networks for Multivariate Time Series with Missing Values |
| title_short | Recurrent Neural Networks for Multivariate Time Series with Missing Values |
| title_sort | recurrent neural networks for multivariate time series with missing values |
| url | https://hdl.handle.net/1721.1/134960 |
| work_keys_str_mv | AT chezhengping recurrentneuralnetworksformultivariatetimeserieswithmissingvalues AT purushothamsanjay recurrentneuralnetworksformultivariatetimeserieswithmissingvalues AT chokyunghyun recurrentneuralnetworksformultivariatetimeserieswithmissingvalues AT sontagdavid recurrentneuralnetworksformultivariatetimeserieswithmissingvalues AT liuyan recurrentneuralnetworksformultivariatetimeserieswithmissingvalues |