Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model
PM<sub>2.5</sub> is one of the main pollutants that cause air pollution, and high concentrations of PM<sub>2.5</sub> seriously threaten human health. Therefore, an accurate prediction of PM<sub>2.5</sub> concentration has great practical significance for air quali...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-09-01
|
| Series: | Atmosphere |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4433/12/9/1211 |
| _version_ | 1797520264219590656 |
|---|---|
| author | Xuchu Jiang Yiwen Luo Biao Zhang |
| author_facet | Xuchu Jiang Yiwen Luo Biao Zhang |
| author_sort | Xuchu Jiang |
| collection | DOAJ |
| description | PM<sub>2.5</sub> is one of the main pollutants that cause air pollution, and high concentrations of PM<sub>2.5</sub> seriously threaten human health. Therefore, an accurate prediction of PM<sub>2.5</sub> concentration has great practical significance for air quality detection, air pollution restoration, and human health. This paper uses the historical air quality concentration data and meteorological data of the Beijing Olympic Sports Center as the research object. This paper establishes a long short-term memory (LSTM) model with a time window size of 12, establishes a T-shape light gradient boosting machine (TSLightGBM) model that uses all information in the time window as the next period of prediction input, and establishes a LSTM-TSLightGBM model pair based on an optimal weighted combination method. PM<sub>2.5</sub> hourly concentration is predicted. The prediction results on the test set show that the mean squared error (MAE), root mean squared error (RMSE), and symmetric mean absolute percentage error (SMAPE) of the LSTM-TSLightGBM model are 11.873, 22.516, and 19.540%, respectively. Compared with LSTM, TSLightGBM, the recurrent neural network (RNN), and other models, the LSTM-TSLightGBM model has a lower MAE, RMSE, and SMAPE, and higher prediction accuracy for PM<sub>2.5</sub> and better goodness-of-fit. |
| first_indexed | 2024-03-10T07:54:19Z |
| format | Article |
| id | doaj.art-876594d5629e42d0ae70f2357836bcbb |
| institution | Directory Open Access Journal |
| issn | 2073-4433 |
| language | English |
| last_indexed | 2024-03-10T07:54:19Z |
| publishDate | 2021-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Atmosphere |
| spelling | doaj.art-876594d5629e42d0ae70f2357836bcbb2023-11-22T12:00:54ZengMDPI AGAtmosphere2073-44332021-09-01129121110.3390/atmos12091211Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination ModelXuchu Jiang0Yiwen Luo1Biao Zhang2School of Statistics and Mathematics, Zhongnan University of Economics and Law, Wuhan 430073, ChinaSchool of Statistics and Mathematics, Zhongnan University of Economics and Law, Wuhan 430073, ChinaSchool of Computer Science, Liaocheng University, Liaocheng 252059, ChinaPM<sub>2.5</sub> is one of the main pollutants that cause air pollution, and high concentrations of PM<sub>2.5</sub> seriously threaten human health. Therefore, an accurate prediction of PM<sub>2.5</sub> concentration has great practical significance for air quality detection, air pollution restoration, and human health. This paper uses the historical air quality concentration data and meteorological data of the Beijing Olympic Sports Center as the research object. This paper establishes a long short-term memory (LSTM) model with a time window size of 12, establishes a T-shape light gradient boosting machine (TSLightGBM) model that uses all information in the time window as the next period of prediction input, and establishes a LSTM-TSLightGBM model pair based on an optimal weighted combination method. PM<sub>2.5</sub> hourly concentration is predicted. The prediction results on the test set show that the mean squared error (MAE), root mean squared error (RMSE), and symmetric mean absolute percentage error (SMAPE) of the LSTM-TSLightGBM model are 11.873, 22.516, and 19.540%, respectively. Compared with LSTM, TSLightGBM, the recurrent neural network (RNN), and other models, the LSTM-TSLightGBM model has a lower MAE, RMSE, and SMAPE, and higher prediction accuracy for PM<sub>2.5</sub> and better goodness-of-fit.https://www.mdpi.com/2073-4433/12/9/1211PM<sub>2.5</sub> concentrationLSTMTSLightGBMtime windowfeature construction |
| spellingShingle | Xuchu Jiang Yiwen Luo Biao Zhang Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model Atmosphere PM<sub>2.5</sub> concentration LSTM TSLightGBM time window feature construction |
| title | Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model |
| title_full | Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model |
| title_fullStr | Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model |
| title_full_unstemmed | Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model |
| title_short | Prediction of PM<sub>2.5</sub> Concentration Based on the LSTM-TSLightGBM Variable Weight Combination Model |
| title_sort | prediction of pm sub 2 5 sub concentration based on the lstm tslightgbm variable weight combination model |
| topic | PM<sub>2.5</sub> concentration LSTM TSLightGBM time window feature construction |
| url | https://www.mdpi.com/2073-4433/12/9/1211 |
| work_keys_str_mv | AT xuchujiang predictionofpmsub25subconcentrationbasedonthelstmtslightgbmvariableweightcombinationmodel AT yiwenluo predictionofpmsub25subconcentrationbasedonthelstmtslightgbmvariableweightcombinationmodel AT biaozhang predictionofpmsub25subconcentrationbasedonthelstmtslightgbmvariableweightcombinationmodel |