Atmospheric particulate matter (PM) has major threats to global health, especially in urban regions around the world. Dhaka, Narayanganj and Gazipur of Bangladesh are positioned as top ranking polluted metropolitan cities in the world. This study assessed the performance of the application of hybrid models, that is, Autoregressive Integrated Moving Average (ARIMA)-Artificial Neural Network (ANN), ARIMA-Support Vector Machine (SVM) and Principle Component Regression (PCR) along with Decision Tree (DT) and CatBoost deep learning model to predict the ambient PM_2.5 concentrations. The data from January 2013 to May 2019 with 2342 observations were utilized in this study. Eighty percent of the data was used as training and the rest of the dataset was employed as testing. The performance of the models was evaluated by R², RMSE and MAE value. Among the models, CatBoost performed best for predicting PM_2.5 for all the stations. The RMSE values during the test period were 12.39 µg m⁻³, 13.06 µg m⁻³ and 12.97 µg m⁻³ for Dhaka, Narayanganj and Gazipur, respectively. Nonetheless, the ARIMA-ANN and DT methods also provided acceptable results. The study suggests adopting deep learning models for predicting atmospheric PM_2.5 in Bangladesh.

中文翻译：

---

ARIMA-ANN，ARIMA-SVM，DT和CatBoost在孟加拉国大气PM2.5预报中的潜力

大气颗粒物（PM）对全球健康构成重大威胁，尤其是在世界范围内的城市地区。孟加拉国的达卡（Dhaka），纳拉扬甘杰（Narayanganj）和加兹普尔（Gazipur）被列为世界上排名最高的污染大都市。这项研究评估了混合模型（即自回归综合移动平均（ARIMA）-人工神经网络（ANN），ARIMA-支持向量机（SVM）和主成分回归（PCR）以及决策树（ DT）和CatBoost深度学习模型来预测环境PM _2.5浓度。本研究利用了2013年1月至2019年5月的2342次观测数据。80％的数据用作训练，其余数据集用作测试。通过R ²，RMSE和MAE值评估模型的性能。在这些模型中，CatBoost在预测所有站点的PM _2.5方面表现最佳。在测试期间的RMSE值分别为12.39微克米^-3，13.06微克米^-3和12.97微克米^-3达卡，纳拉扬甘杰市和Gazipur，分别。尽管如此，ARIMA-ANN和DT方法也提供了可接受的结果。研究建议采用深度学习模型来预测孟加拉国的大气PM _2.5。