Exploration of Water Level Forecast for Hongze Lake Based on Bayesian Model Averaging
YANG Changwen
WANG Chao
LEI Xiaohui
XU Ke
Abstract：The Bayesian model averaging method provides a statistical framework for evaluating and comparing multiple candidate models.It provides more accurate and robust prediction and inference results by combining the prediction results of multiple models and estimating their weights.The water level prediction model for Hongze Lake was established in this article by using recurrent neural networks such as Long Short Term Memory Network(LSTM),Elman Network(Elman),and Control Recurrent Unit(GRU),and the prediction results of these three models were combined and verified by BMA method.The results showed that the BMA combination model based on Bayesian combination method had higher prediction accuracy than the single model,and improved the stability of prediction.
Keywords：Bayesian model averagingHongze Lakewater level forecastmodel collection
Publication Date：2025-01-27
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 80-86 )