| 電子情報通信学会 研究会発表申込システム 講演論文 詳細 |
技報閲覧サービス [ログイン] 技報アーカイブ |
| トップに戻る | 前のページに戻る | [Japanese] / [English] |
| 講演抄録／キーワード | ||
| 講演名 | 2024-01-25 16:40 ［ポスター講演］Dust Forecasting using ARIMA and ARIMAX Models ○Nuntouchaporn Prateepausanont（College of digital science, Prince of Songkla Univ.）・Anny Mardjo（Prince of Songkla Univ. , Trang Campus）・Chidchanok Choksuchat・Korakot Wichitsa-nguan Jetwanna・Thanathip Limna（Prince of Songkla Univ.）  IA2023-69 |
|
| 抄録 | （和） | Southern Thailand faces air pollution problems from transboundary haze, as shown in the Pollution Control Department, caused by dust particles moving from areas within Indonesia. This haze affects the health of the population in the southern region over a wide area. Therefore, predicting the amount of dust that will cover the area would enable us to promptly propose solutions to reduce the impact of dust. the environmental science and internet architecture/applications can intersect in various ways, particularly concerning data collection, data sharing, sustainability, and the application of internet technologies for environmental monitoring and research. Collaboration between experts from both fields can lead to innovative solutions for addressing environmental challenges and reducing the environmental footprint of internet technologies. This research presents dust forecasting in Southern Thailand by considering the utilization of ARIMA for univariate time series forecasting, ARIMAX for multivariate time series forecasting with exogenous variables. The study uses dataset that generated from the RCNAME for the years spanning from January 2016 to December 2022. The remote sensors, including PCD and TMD, were collected concurrently with the laboratory data. The research methodology consists of data cleansing, exploratory data analysis, and fitting time-series models. Finally, the ARIMA (0,1,1) model offered the highest performance model for forecasting due to the lowest MAE (=1.6857), MAPE (=3.76126) and RMSE (=0.2304), while the ARIMAX (0,1,2) model has a lower MAE (=1.6857), MAPE (=27.30) and RMSE (=2.2937). The highest performance model for forecasting is the ARIMA (0,1,1) model. |
| （英） | Southern Thailand faces air pollution problems from transboundary haze, as shown in the Pollution Control Department, caused by dust particles moving from areas within Indonesia. This haze affects the health of the population in the southern region over a wide area. Therefore, predicting the amount of dust that will cover the area would enable us to promptly propose solutions to reduce the impact of dust. the environmental science and internet architecture/applications can intersect in various ways, particularly concerning data collection, data sharing, sustainability, and the application of internet technologies for environmental monitoring and research. Collaboration between experts from both fields can lead to innovative solutions for addressing environmental challenges and reducing the environmental footprint of internet technologies. This research presents dust forecasting in Southern Thailand by considering the utilization of ARIMA for univariate time series forecasting, ARIMAX for multivariate time series forecasting with exogenous variables. The study uses dataset that generated from the RCNAME for the years spanning from January 2016 to December 2022. The remote sensors, including PCD and TMD, were collected concurrently with the laboratory data. The research methodology consists of data cleansing, exploratory data analysis, and fitting time-series models. Finally, the ARIMA (0,1,1) model offered the highest performance model for forecasting due to the lowest MAE (=1.6857), MAPE (=3.76126) and RMSE (=0.2304), while the ARIMAX (0,1,2) model has a lower MAE (=1.6857), MAPE (=27.30) and RMSE (=2.2937). The highest performance model for forecasting is the ARIMA (0,1,1) model. | |
| キーワード | （和） | PM / ARIMA / ARIMAX / Forecasting / / / / |
| （英） | PM / ARIMA / ARIMAX / Forecasting / / / / | |
| 文献情報 | 信学技報, vol. 123, no. 372, IA2023-69, pp. 52-57, 2024年1月. | |
| 資料番号 | IA2023-69 | |
| 発行日 | 2024-01-18 (IA) | |
| ISSN | Online edition: ISSN 2432-6380 | |
| 著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します．(許諾番号：10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) | |
| PDFダウンロード | IA2023-69 |
|
[研究会発表申込システムのトップページに戻る]