PR-FCNN: a data-driven hybrid approach for predicting PM2.5 concentration DOI Creative Commons
Syed Azeem Inam, Abdullah Ayub Khan, Tehseen Mazhar

и другие.

Discover Artificial Intelligence, Год журнала: 2024, Номер 4(1)

Опубликована: Ноя. 3, 2024

The atmosphere's fine articulate Matter (PM2.5) poses various health-related risks. Even though multiple efforts have been made to lower the emissions of these substances, mortality rate is continuously increasing, requiring immediate inclination scientific community towards design and development advanced predictive models. Conventional statistical approaches become dormant due their limitations in capturing innate relationships between pollutants, particularly for predicting PM2.5 concentrations. In contrast, machine deep learning techniques shown great potential forecasting air quality, providing more accuracy than its predecessor techniques. present study investigates utilization hybrid by integrating models with improve prediction capabilities concentration. It uses datasets from World Air Quality Index (WAQI) State Global (SOGA) analyze performance on both daily annual data, respectively. This ensures model's effectiveness a diversified dataset. implements Random Forest (RF), Polynomial Regression (PR), XGBoost, Extra Tree Regressor (ETR) coupled Fully Connected Neural Network (FCNN), Long Short-Term Memory (LSTM), Bi-directional LSTM (Bi-LSTM) obtaining optimized results. Finally, after thorough investigation, PR model FCNN (PR-FCNN) found be best improved R-squared (R2) values, portraying concentration accurately. Based experimentation, preset recommends implementing approaches, offering better especially PM2.5.

Язык: Английский

Comparative Analysis of Multiple Deep Learning Models for Forecasting Monthly Ambient PM2.5 Concentrations: A Case Study in Dezhou City, China DOI Creative Commons

Zhenfang He,

Qingchun Guo

Atmosphere, Год журнала: 2024, Номер 15(12), С. 1432 - 1432

Опубликована: Ноя. 28, 2024

Ambient air pollution affects human health, vegetative growth and sustainable socio-economic development. Therefore, data in Dezhou City China are collected from January 2014 to December 2023, multiple deep learning models used forecast PM2.5 concentrations. The ability of the is evaluated compared with observed using various statistical parameters. Although all eight can accomplish forecasting assignments, precision accuracy CNN-GRU-LSTM method 34.28% higher than that ANN method. result shows has best performance other seven models, achieving an R (correlation coefficient) 0.9686 RMSE (root mean square error) 4.6491 μg/m3. values CNN, GRU LSTM 57.00%, 35.98% 32.78% method, respectively. results reveal predictor remarkably improves performances benchmark overall forecasting. This research provides a new perspective for predictive ambient model provide scientific basis prevention control.

Язык: Английский

Процитировано

17
A nonlinear grey model with seasonal weighted fractional accumulation for triangular fuzzy number series and its application to forecast PM2.5 DOI
Zhenxiu Cao, Xiangyan Zeng, Shuli Yan

и другие.

Grey Systems Theory and Application, Год журнала: 2025, Номер unknown

Опубликована: Апрель 25, 2025

Purpose Accurate prediction of PM2.5 concentration is essential for the government to formulate and implement effective environmental policies management measures improve air quality. series exhibits seasonal, nonlinear, uncertain characteristics. A seasonal weighted fractional nonlinear grey model triangular fuzzy number established based on Bernoulli by introducing accumulation generating operator. Design/methodology/approach First, actual sequence processed using a new operator weaken its seasonality. The sine function time power are introduced into perform processing again, thereby enhancing model’s adaptability series. Secondly, parameters transformed matrix form so as directly Additionally, optimal algorithm selected through comparison experiments used determine parameters. Findings Five models predict concentrations in Shanghai, China San Francisco, United States America (USA). findings show that with operator, can better simulate characteristics compared other models. Then, next four quarters two cities predicted analyzed. Originality/value dynamic volatility. When represented series, it reflects complexity uncertainty data, which helps people make more accurate decisions. capacity precisely forecast improved large part this work.

Язык: Английский

Процитировано

0
PR-FCNN: a data-driven hybrid approach for predicting PM2.5 concentration DOI Creative Commons
Syed Azeem Inam, Abdullah Ayub Khan, Tehseen Mazhar

и другие.

Discover Artificial Intelligence, Год журнала: 2024, Номер 4(1)

Опубликована: Ноя. 3, 2024

The atmosphere's fine articulate Matter (PM2.5) poses various health-related risks. Even though multiple efforts have been made to lower the emissions of these substances, mortality rate is continuously increasing, requiring immediate inclination scientific community towards design and development advanced predictive models. Conventional statistical approaches become dormant due their limitations in capturing innate relationships between pollutants, particularly for predicting PM2.5 concentrations. In contrast, machine deep learning techniques shown great potential forecasting air quality, providing more accuracy than its predecessor techniques. present study investigates utilization hybrid by integrating models with improve prediction capabilities concentration. It uses datasets from World Air Quality Index (WAQI) State Global (SOGA) analyze performance on both daily annual data, respectively. This ensures model's effectiveness a diversified dataset. implements Random Forest (RF), Polynomial Regression (PR), XGBoost, Extra Tree Regressor (ETR) coupled Fully Connected Neural Network (FCNN), Long Short-Term Memory (LSTM), Bi-directional LSTM (Bi-LSTM) obtaining optimized results. Finally, after thorough investigation, PR model FCNN (PR-FCNN) found be best improved R-squared (R2) values, portraying concentration accurately. Based experimentation, preset recommends implementing approaches, offering better especially PM2.5.

Язык: Английский

Процитировано

1