File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/TBDATA.2020.3005368
- Scopus: eid_2-s2.0-85089760243
- WOS: WOS:000822368700001
Supplementary
- Citations:
- Appears in Collections:
Article: A Domain-Specific Bayesian Deep-Learning Approach for Air Pollution Forecast
Title | A Domain-Specific Bayesian Deep-Learning Approach for Air Pollution Forecast |
---|---|
Authors | |
Keywords | Air pollution forecast Bayesian deep-learning Domain-specific knowledge Prediction fusion Prediction uncertainty |
Issue Date | 2022 |
Citation | IEEE Transactions on Big Data, 2022, v. 8, n. 4, p. 1034-1046 How to Cite? |
Abstract | Predicting air pollution concentration is crucial and beneficial for public health. This study proposes a domain-specific Bayesian deep-learning model for long-term air pollution forecast in China and the United Kingdom. Our proposed model carries three novelties: First, a domain-specific knowledge is integrated to take into account the strong statistical relationship between PM 2.5 and PM 10 as a regularization term; Second, an attention layer is included to capture the influential historical feature and the recursive temporal correlation of air quality data; Third, results generated from different multi-step forecast strategies are combined based on corresponding uncertainty measures to improve our model's performance. Our model outperforms other baseline models. Results show that incorporating Bayesian and domain-specific knowledge into the deep learning model can reduce the prediction errors by a maximum of 3.7% and 12.4%, for Beijing and London, respectively. Specifically, incorporating domain-specific knowledge into the Bayesian deep-learning model reduces prediction errors whilst the integration of Bayesian techniques allows the fusion of different forecast strategies to improve prediction accuracy. In future, additional influential domain-specific features can be added to further improve our deep-learning model's prediction accuracy and interpretability. |
Persistent Identifier | http://hdl.handle.net/10722/336797 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Han, Yang | - |
dc.contributor.author | Lam, Jacqueline C.K. | - |
dc.contributor.author | Li, Victor O.K. | - |
dc.contributor.author | Zhang, Qi | - |
dc.date.accessioned | 2024-02-29T06:56:36Z | - |
dc.date.available | 2024-02-29T06:56:36Z | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | IEEE Transactions on Big Data, 2022, v. 8, n. 4, p. 1034-1046 | - |
dc.identifier.uri | http://hdl.handle.net/10722/336797 | - |
dc.description.abstract | Predicting air pollution concentration is crucial and beneficial for public health. This study proposes a domain-specific Bayesian deep-learning model for long-term air pollution forecast in China and the United Kingdom. Our proposed model carries three novelties: First, a domain-specific knowledge is integrated to take into account the strong statistical relationship between PM 2.5 and PM 10 as a regularization term; Second, an attention layer is included to capture the influential historical feature and the recursive temporal correlation of air quality data; Third, results generated from different multi-step forecast strategies are combined based on corresponding uncertainty measures to improve our model's performance. Our model outperforms other baseline models. Results show that incorporating Bayesian and domain-specific knowledge into the deep learning model can reduce the prediction errors by a maximum of 3.7% and 12.4%, for Beijing and London, respectively. Specifically, incorporating domain-specific knowledge into the Bayesian deep-learning model reduces prediction errors whilst the integration of Bayesian techniques allows the fusion of different forecast strategies to improve prediction accuracy. In future, additional influential domain-specific features can be added to further improve our deep-learning model's prediction accuracy and interpretability. | - |
dc.language | eng | - |
dc.relation.ispartof | IEEE Transactions on Big Data | - |
dc.subject | Air pollution forecast | - |
dc.subject | Bayesian deep-learning | - |
dc.subject | Domain-specific knowledge | - |
dc.subject | Prediction fusion | - |
dc.subject | Prediction uncertainty | - |
dc.title | A Domain-Specific Bayesian Deep-Learning Approach for Air Pollution Forecast | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/TBDATA.2020.3005368 | - |
dc.identifier.scopus | eid_2-s2.0-85089760243 | - |
dc.identifier.volume | 8 | - |
dc.identifier.issue | 4 | - |
dc.identifier.spage | 1034 | - |
dc.identifier.epage | 1046 | - |
dc.identifier.eissn | 2332-7790 | - |
dc.identifier.isi | WOS:000822368700001 | - |