Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks
Time series classification and forecasting have long been studied with the traditional statistical methods. Recently, deep learning achieved remarkable successes in areas such as image, text, video, audio processing, etc. However, research studies conducted with deep neural networks in these fields...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-12-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/24/7211 |
| _version_ | 1797544479182290944 |
|---|---|
| author | Kun Zhou Wenyong Wang Teng Hu Kai Deng |
| author_facet | Kun Zhou Wenyong Wang Teng Hu Kai Deng |
| author_sort | Kun Zhou |
| collection | DOAJ |
| description | Time series classification and forecasting have long been studied with the traditional statistical methods. Recently, deep learning achieved remarkable successes in areas such as image, text, video, audio processing, etc. However, research studies conducted with deep neural networks in these fields are not abundant. Therefore, in this paper, we aim to propose and evaluate several state-of-the-art neural network models in these fields. We first review the basics of representative models, namely long short-term memory and its variants, the temporal convolutional network and the generative adversarial network. Then, long short-term memory with autoencoder and attention-based models, the temporal convolutional network and the generative adversarial model are proposed and applied to time series classification and forecasting. Gaussian sliding window weights are proposed to speed the training process up. Finally, the performances of the proposed methods are assessed using five optimizers and loss functions with the public benchmark datasets, and comparisons between the proposed temporal convolutional network and several classical models are conducted. Experiments show the proposed models’ effectiveness and confirm that the temporal convolutional network is superior to long short-term memory models in sequence modeling. We conclude that the proposed temporal convolutional network reduces time consumption to around 80% compared to others while retaining the same accuracy. The unstable training process for generative adversarial network is circumvented by tuning hyperparameters and carefully choosing the appropriate optimizer of “Adam”. The proposed generative adversarial network also achieves comparable forecasting accuracy with traditional methods. |
| first_indexed | 2024-03-10T14:01:05Z |
| format | Article |
| id | doaj.art-737700bee2b84c9ab1042e9ad61e78fe |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T14:01:05Z |
| publishDate | 2020-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-737700bee2b84c9ab1042e9ad61e78fe2023-11-21T01:05:15ZengMDPI AGSensors1424-82202020-12-012024721110.3390/s20247211Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial NetworksKun Zhou0Wenyong Wang1Teng Hu2Kai Deng3School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, ChinaSchool of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, ChinaSchool of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, ChinaInstitute of Computer Application, China Academy of Engineering Physics, Mianyang 621900, ChinaTime series classification and forecasting have long been studied with the traditional statistical methods. Recently, deep learning achieved remarkable successes in areas such as image, text, video, audio processing, etc. However, research studies conducted with deep neural networks in these fields are not abundant. Therefore, in this paper, we aim to propose and evaluate several state-of-the-art neural network models in these fields. We first review the basics of representative models, namely long short-term memory and its variants, the temporal convolutional network and the generative adversarial network. Then, long short-term memory with autoencoder and attention-based models, the temporal convolutional network and the generative adversarial model are proposed and applied to time series classification and forecasting. Gaussian sliding window weights are proposed to speed the training process up. Finally, the performances of the proposed methods are assessed using five optimizers and loss functions with the public benchmark datasets, and comparisons between the proposed temporal convolutional network and several classical models are conducted. Experiments show the proposed models’ effectiveness and confirm that the temporal convolutional network is superior to long short-term memory models in sequence modeling. We conclude that the proposed temporal convolutional network reduces time consumption to around 80% compared to others while retaining the same accuracy. The unstable training process for generative adversarial network is circumvented by tuning hyperparameters and carefully choosing the appropriate optimizer of “Adam”. The proposed generative adversarial network also achieves comparable forecasting accuracy with traditional methods.https://www.mdpi.com/1424-8220/20/24/7211time series forecastingtime series classificationlong short-term memoryattention mechanismgenerative adversarial network |
| spellingShingle | Kun Zhou Wenyong Wang Teng Hu Kai Deng Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks Sensors time series forecasting time series classification long short-term memory attention mechanism generative adversarial network |
| title | Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks |
| title_full | Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks |
| title_fullStr | Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks |
| title_full_unstemmed | Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks |
| title_short | Time Series Forecasting and Classification Models Based on Recurrent with Attention Mechanism and Generative Adversarial Networks |
| title_sort | time series forecasting and classification models based on recurrent with attention mechanism and generative adversarial networks |
| topic | time series forecasting time series classification long short-term memory attention mechanism generative adversarial network |
| url | https://www.mdpi.com/1424-8220/20/24/7211 |
| work_keys_str_mv | AT kunzhou timeseriesforecastingandclassificationmodelsbasedonrecurrentwithattentionmechanismandgenerativeadversarialnetworks AT wenyongwang timeseriesforecastingandclassificationmodelsbasedonrecurrentwithattentionmechanismandgenerativeadversarialnetworks AT tenghu timeseriesforecastingandclassificationmodelsbasedonrecurrentwithattentionmechanismandgenerativeadversarialnetworks AT kaideng timeseriesforecastingandclassificationmodelsbasedonrecurrentwithattentionmechanismandgenerativeadversarialnetworks |