When Nitrogen Dioxide (NO ₂ ) and Sulfur Dioxide (SO ₂ ) mix, they cause pulmonary fibrosis, and severe public health issues. Therefore, introducing deep learning models to predict changes in Nitrogen Dioxide and Sulfur Dioxide air pollution can enable the early formulation of air pollution policies. The study proposed an effective method for predicting air pollution levels in Taiwan by utilizing the Kalman filtering technique and employing the seasonal gated recurrent units (SGRU) deep learning model. The data used in this study were obtained from the Environmental Protection Administration of Taiwan. These data include Nitrogen Dioxide and Sulfur Dioxide air pollution measurements obtained between 2005 and 2021 from various monitoring stations in Tai-wan. The proposed SGRU model was compared with six other commonly used prediction methods, namely autoregressive integrated moving average (ARIMA), seasonal autoregressive integrated moving average (SARIMA), exponential smooth-ing (ETS), Holt-Winters exponential smoothing (HWETS), support vector regression (SVR), and seasonal long short-term memory (SLSTM). The mean absolute percentage error (MAPE) and root mean square error (RMSE) were used to evaluate the accuracy of the models. The study results revealed that the SGRU model achieved the lowest MAPE value of 0.006 for nitrogen dioxide in Shanhua, the highest value of 1.08 for Keelung, and the lowest value of 0.005 for sulfur dioxide in Taitung, with the highest value of 2.47 for Cianjhen. This demonstrates the broad applicability of the SGRU model to the Taiwan region. This study shows that the SGRU model generally applies in Taiwan and has the lowest prediction error. The SGRU model can predict air pollution levels more accurately, and this precise forecasting is crucial for governments, public health agencies, and medical institutions to develop preventive and response measures. It helps protect plants, animals, aquatic life, and the ecological environment while maintaining ecological balance. Incorporating air pollution predictions into urban planning allows for early infrastructure and green energy planning, thereby reducing air pollution levels and minimizing the resulting ecosystem degradation, promoting a healthier urban living environment.

According to the World Energy Investment 2018 report, the global annual investment in renewable energy exceeded USD 200 billion for eight consecutive years until 2017. In this paper, a deep-learning-based time-series prediction method, namely a gated recurrent unit (GRU)-based prediction method, is proposed to predict energy generation in Taiwan. Data on thermal power (coal, oil, and gas power), renewable energy (conventional hydropower, solar power, and wind power), pumped hydropower, and nuclear power generation for 1991 to 2020 were obtained from the Bureau of Energy, Ministry of Economic Affairs, Taiwan, and the Taiwan Power Company. The proposed GRU-based method was compared with six common forecasting methods: autoregressive integrated moving average, exponential smoothing (ETS), Holt–Winters ETS, support vector regression (SVR), whale-optimization-algorithm-based SVR, and long short-term memory. Among the methods compared, the proposed method had the lowest mean absolute percentage error and root mean square error and thus the highest accuracy. Government agencies and power companies in Taiwan can use the predictions of accurate energy forecasting models as references to formulate energy policies and design plans for the development of alternative energy sources.