A Clustering-Based Deep Learning Method for Water Level Prediction

Abstract

Accurate water level prediction systems improve safety and quality of life. This study introduces a method that uses clustering and deep learning of multisite data to enhance the water level prediction of the Three Gorges Dam. The results show that Cluster-GRU-based can provide accurate forecasts for up to seven days.