Applying iTransformer for Saigon River Water Level Forecasting

Water level forecasting holds immense significance for the effective management of water resources with particular relevance to the Saigon River. As a vital waterway flowing through Ho Chi Minh City, the Saigon River plays a crucial role in supporting various aspects of urban life including agriculture, transportation, and recreation. Accurate forecasting of water levels in the Saigon River is essential for mitigating the risks of flooding, which is a recurrent challenge faced by the city. Moreover, with the increasing urbanization and industrialization in the region, precise prediction of water levels in the Saigon River is vital for sustainable development, ensuring adequate water supply for domestic, industrial, and agricultural needs while preserving the ecological balance of the river ecosystem. Recently, along with the advancement of machine learning and artificial intelligence algorithms, numerous studies have been conducted and achieved promising results in river water level forecasting. Most successful research directions focus on applying time series analysis algorithms such as ARIMA [1], [6], XGBOOST [2], [7], and LSTM [3], [8], [9]. In this study, we propose a solution using iTransformer [11], which achieves superior accuracy compared to traditional methods.