Author(s): Ling Dai; Shaolin Li; Guangbiao Liu; Chuanbo Ji
Linked Author(s):
Keywords: Dam deformation; Cluster; Machine learning; Bayesian optimization
Abstract: To improve the accuracy of dam deformation prediction models, the dam deformation prediction model based on different machine learning algorithms on different categories of data generated by different clustering methods was proposed, considering the inconsistency of monitoring data in different parts and stages of the project. Simultaneously, Bayesian optimization algorithm was introduced to optimize model hyperparameter. Moreover, the optimal model combination from the clustering algorithm models and machine learning models was selected based on the principle of minimizing root mean square error. The research results show that the prediction accuracy of the proposed model was better than the model directly using all data. Furthermore, the proposed model has strong scalability and can adapt to more types of data by introducing more clustering algorithms and machine learning algorithms, thereby improving the prediction accuracy and applicability of the model.
Year: 2024