A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting
Time series forecasting has attracted wide attention in recent decades. However, some time series are imbalanced and show different patterns between special and normal periods, leading to the prediction accuracy degradation of special periods. In this paper, we aim to develop a unified model to alle...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tsinghua University Press
2021-12-01
|
| Series: | Big Data Mining and Analytics |
| Subjects: | |
| Online Access: | https://www.sciopen.com/article/10.26599/BDMA.2021.9020011 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832572795536539648 |
|---|---|
| author | Chenyu Hou Jiawei Wu Bin Cao Jing Fan |
| author_facet | Chenyu Hou Jiawei Wu Bin Cao Jing Fan |
| author_sort | Chenyu Hou |
| collection | DOAJ |
| description | Time series forecasting has attracted wide attention in recent decades. However, some time series are imbalanced and show different patterns between special and normal periods, leading to the prediction accuracy degradation of special periods. In this paper, we aim to develop a unified model to alleviate the imbalance and thus improving the prediction accuracy for special periods. This task is challenging because of two reasons: (1) the temporal dependency of series, and (2) the tradeoff between mining similar patterns and distinguishing different distributions between different periods. To tackle these issues, we propose a self-attention-based time-varying prediction model with a two-stage training strategy. First, we use an encoder-decoder module with the multi-head self-attention mechanism to extract common patterns of time series. Then, we propose a time-varying optimization module to optimize the results of special periods and eliminate the imbalance. Moreover, we propose reverse distance attention in place of traditional dot attention to highlight the importance of similar historical values to forecast results. Finally, extensive experiments show that our model performs better than other baselines in terms of mean absolute error and mean absolute percentage error. |
| format | Article |
| id | doaj-art-2da2db05295543008c6b24212953e52a |
| institution | Kabale University |
| issn | 2096-0654 |
| language | English |
| publishDate | 2021-12-01 |
| publisher | Tsinghua University Press |
| record_format | Article |
| series | Big Data Mining and Analytics |
| spelling | doaj-art-2da2db05295543008c6b24212953e52a2025-02-02T06:50:33ZengTsinghua University PressBig Data Mining and Analytics2096-06542021-12-014426627810.26599/BDMA.2021.9020011A Deep-Learning Prediction Model for Imbalanced Time Series Data ForecastingChenyu Hou0Jiawei Wu1Bin Cao2Jing Fan3<institution>College of Computer Science and Technology, Zhejiang University of Technology</institution>, <city>Hangzhou</city> <postal-code>310023</postal-code>, <country>China</country><institution>College of Computer Science and Technology, Zhejiang University of Technology</institution>, <city>Hangzhou</city> <postal-code>310023</postal-code>, <country>China</country><institution>College of Computer Science and Technology, Zhejiang University of Technology</institution>, <city>Hangzhou</city> <postal-code>310023</postal-code>, <country>China</country><institution>College of Computer Science and Technology, Zhejiang University of Technology</institution>, <city>Hangzhou</city> <postal-code>310023</postal-code>, <country>China</country>Time series forecasting has attracted wide attention in recent decades. However, some time series are imbalanced and show different patterns between special and normal periods, leading to the prediction accuracy degradation of special periods. In this paper, we aim to develop a unified model to alleviate the imbalance and thus improving the prediction accuracy for special periods. This task is challenging because of two reasons: (1) the temporal dependency of series, and (2) the tradeoff between mining similar patterns and distinguishing different distributions between different periods. To tackle these issues, we propose a self-attention-based time-varying prediction model with a two-stage training strategy. First, we use an encoder-decoder module with the multi-head self-attention mechanism to extract common patterns of time series. Then, we propose a time-varying optimization module to optimize the results of special periods and eliminate the imbalance. Moreover, we propose reverse distance attention in place of traditional dot attention to highlight the importance of similar historical values to forecast results. Finally, extensive experiments show that our model performs better than other baselines in terms of mean absolute error and mean absolute percentage error.https://www.sciopen.com/article/10.26599/BDMA.2021.9020011time series forecastingimbalanced datadeep learningprediction model |
| spellingShingle | Chenyu Hou Jiawei Wu Bin Cao Jing Fan A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting Big Data Mining and Analytics time series forecasting imbalanced data deep learning prediction model |
| title | A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting |
| title_full | A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting |
| title_fullStr | A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting |
| title_full_unstemmed | A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting |
| title_short | A Deep-Learning Prediction Model for Imbalanced Time Series Data Forecasting |
| title_sort | deep learning prediction model for imbalanced time series data forecasting |
| topic | time series forecasting imbalanced data deep learning prediction model |
| url | https://www.sciopen.com/article/10.26599/BDMA.2021.9020011 |
| work_keys_str_mv | AT chenyuhou adeeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT jiaweiwu adeeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT bincao adeeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT jingfan adeeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT chenyuhou deeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT jiaweiwu deeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT bincao deeplearningpredictionmodelforimbalancedtimeseriesdataforecasting AT jingfan deeplearningpredictionmodelforimbalancedtimeseriesdataforecasting |