Short-term forecasting of PM10and PM2.5 concentrations with Facebook's Prophet Model at the Belgrade-Zeleno brdo
Чланак у часопису (Рецензирана верзија)
Метаподаци
Приказ свих података о документуАпстракт
We demonstrate the use of Facebook's Prophet (usually just called Prophet) model for short-term air quality forecasting at Belgrade-Zeleno brdo monitoring station. To address missing data, we applied minimally-altering data distribution imputation techniques. Linear interpolation proved effective for short-term gaps (1–3 hours), hourly mean method for mid-term gaps (24–26 hours), and Hermite interpolation polynomial for long-term gaps (132–148 hours). The most significant data change was a 3.4% shift in skewness. Partitioning the time series enabled a detailed quality assessment of the Prophet model, with PM2.5predictions being more precise than PM10. Using the longest time series for forecasting yielded absolute errors of 6.5μg/m3forPM10and 2.7μg/m3for PM2.5. Based on 173 forecasts, we anticipate Prophet model root-mean-square values under 6.26μg/m3and 9.99μg/m3for PM2.5 and PM10in 50% of cases. The Prophet model demonst...rates several advantages and yields satisfactory results. In future research, the results obtained from the Prophet model will serve as benchmark values for other models. Additionally, the Prophet model is capable of providing satisfactory air quality forecasting results and will be utilized in future researc
Кључне речи:
time series analysis / time series forecasting / air quality / suspended particlesИзвор:
Geofizika, 2023, 40Институција/група
GraFarTY - JOUR AU - Arnaut, Filip AU - Cvetkov, Vesna AU - Đurić, Dragana AU - Samardžić-Petrović, Mileva PY - 2023 UR - https://grafar.grf.bg.ac.rs/handle/123456789/3275 AB - We demonstrate the use of Facebook's Prophet (usually just called Prophet) model for short-term air quality forecasting at Belgrade-Zeleno brdo monitoring station. To address missing data, we applied minimally-altering data distribution imputation techniques. Linear interpolation proved effective for short-term gaps (1–3 hours), hourly mean method for mid-term gaps (24–26 hours), and Hermite interpolation polynomial for long-term gaps (132–148 hours). The most significant data change was a 3.4% shift in skewness. Partitioning the time series enabled a detailed quality assessment of the Prophet model, with PM2.5predictions being more precise than PM10. Using the longest time series for forecasting yielded absolute errors of 6.5μg/m3forPM10and 2.7μg/m3for PM2.5. Based on 173 forecasts, we anticipate Prophet model root-mean-square values under 6.26μg/m3and 9.99μg/m3for PM2.5 and PM10in 50% of cases. The Prophet model demonstrates several advantages and yields satisfactory results. In future research, the results obtained from the Prophet model will serve as benchmark values for other models. Additionally, the Prophet model is capable of providing satisfactory air quality forecasting results and will be utilized in future researc T2 - Geofizika T1 - Short-term forecasting of PM10and PM2.5 concentrations with Facebook's Prophet Model at the Belgrade-Zeleno brdo VL - 40 DO - 10.15233/gfz.2023.40.7 ER -
@article{ author = "Arnaut, Filip and Cvetkov, Vesna and Đurić, Dragana and Samardžić-Petrović, Mileva", year = "2023", abstract = "We demonstrate the use of Facebook's Prophet (usually just called Prophet) model for short-term air quality forecasting at Belgrade-Zeleno brdo monitoring station. To address missing data, we applied minimally-altering data distribution imputation techniques. Linear interpolation proved effective for short-term gaps (1–3 hours), hourly mean method for mid-term gaps (24–26 hours), and Hermite interpolation polynomial for long-term gaps (132–148 hours). The most significant data change was a 3.4% shift in skewness. Partitioning the time series enabled a detailed quality assessment of the Prophet model, with PM2.5predictions being more precise than PM10. Using the longest time series for forecasting yielded absolute errors of 6.5μg/m3forPM10and 2.7μg/m3for PM2.5. Based on 173 forecasts, we anticipate Prophet model root-mean-square values under 6.26μg/m3and 9.99μg/m3for PM2.5 and PM10in 50% of cases. The Prophet model demonstrates several advantages and yields satisfactory results. In future research, the results obtained from the Prophet model will serve as benchmark values for other models. Additionally, the Prophet model is capable of providing satisfactory air quality forecasting results and will be utilized in future researc", journal = "Geofizika", title = "Short-term forecasting of PM10and PM2.5 concentrations with Facebook's Prophet Model at the Belgrade-Zeleno brdo", volume = "40", doi = "10.15233/gfz.2023.40.7" }
Arnaut, F., Cvetkov, V., Đurić, D.,& Samardžić-Petrović, M.. (2023). Short-term forecasting of PM10and PM2.5 concentrations with Facebook's Prophet Model at the Belgrade-Zeleno brdo. in Geofizika, 40. https://doi.org/10.15233/gfz.2023.40.7
Arnaut F, Cvetkov V, Đurić D, Samardžić-Petrović M. Short-term forecasting of PM10and PM2.5 concentrations with Facebook's Prophet Model at the Belgrade-Zeleno brdo. in Geofizika. 2023;40. doi:10.15233/gfz.2023.40.7 .
Arnaut, Filip, Cvetkov, Vesna, Đurić, Dragana, Samardžić-Petrović, Mileva, "Short-term forecasting of PM10and PM2.5 concentrations with Facebook's Prophet Model at the Belgrade-Zeleno brdo" in Geofizika, 40 (2023), https://doi.org/10.15233/gfz.2023.40.7 . .