Can I read Rapid prediction of urban flood based on disaster-breeding environment clustering and Bayesian optimized deep learning model in the coastal city on EtoBox?

Rapid prediction of urban flood based on disaster-breeding environment clustering and Bayesian optimized deep learning model in the coastal city by Huiliang Wang; Shanlun Xu; Hongshi Xu; Zening Wu; Tianye Wang; Chao Ma is a Environmental Science article available to read on EtoBox.

What is Rapid prediction of urban flood based on disaster-breeding environment clustering and Bayesian optimized deep learning model in the coastal city about?

Rapid prediction of urban flood is essential for sustainable city and society development. The data-driven deep learning model is commonly adopted for flood prediction, but it rarely considers the physical causes of flooding. In the study, we proposed an urban flood prediction model based on K-means cluster and Bayesian optimized deep learning algorithm. The K-means cluster algorithm was utilized for classifying the study area into different flooding cause clusters based on disaster-bearing environment factors. The gradient boosting decision tree (GBDT) algorithm was adopted to establish the prediction model for the areas with the same flooding cause. Furthermore, to improve the model accuracy, the study explored Bayesian algorithm to optimize GBDT model's hyperparameters. The proposed model was verified in the area of Haikou, China. Elevation, slope, density of pipe network, distance to the river and curvature were selected as the disaster-bearing environment factors, and the study area was classified into 3 clusters using K-means cluster based on these factors. Different GBDT models were established for each cluster, then Bayesian algorithm was used to optimize the six hyperparam

Who reads Rapid prediction of urban flood based on disaster-breeding environment clustering and Bayesian optimized deep learning model in the coastal city?

Author

Huiliang Wang; Shanlun Xu; Hongshi Xu; Zening Wu; Tianye Wang; Chao Ma

Publisher

Elsevier BV

Published

2023

Language

EN

Field

Environmental Science (Physical Sciences)