| Abstract |
Weather forecasting is crucial for effective resource management and infrastructure planning in smart cities, necessitating accurate predictions of future weather conditions. However, the inherently complex nature of weather phenomena, including nonlinear parameters such as rainfall, humidity, wind speed, and temperature, poses significant challenges to forecasting efforts. This difficulty is compounded by the unpredictable nature of weather changes. To address this, various time series forecasting models, including ARIMA, SARIMA, Prophet, and Triple Exponential Smoothing models, are employed to develop a reliable weather forecasting system tailored to smart city environments. By utilizing historical weather data and patterns obtained from publicly available datasets like Kaggle, these models strive to improve the precision of weather forecasts by employing evaluation criteria such as mean absolute error, mean squared error, mean absolute percent error, and root mean square error. The outcomes of this research not only promise improved weather predictions for smart cities but also offer potential applications in tourism, natural resource management, agriculture, and smart city infrastructure development. © 2024 selection and editorial matter, Sujit Kumar Pradhan, Srinivas Sethi, Mufti Mahmud; individual chapters, the contributors. |
