Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area
In recent years, the major geological hazard of landslides has greatly impact normal human life. Deep belief networks (DBN) is a commonly used deep learning model, and the DBN hyperparameter determination problem is the key to its application. To improve the accuracy of regional landslide susceptibi...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-08-01
|
| Series: | Frontiers in Environmental Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fenvs.2022.912523/full |
| _version_ | 1817998432808206336 |
|---|---|
| author | Siying Wang Xiaokun Lin Xiaokun Lin Xiaokun Lin Xing Qi Hongde Li Jingjing Yang |
| author_facet | Siying Wang Xiaokun Lin Xiaokun Lin Xiaokun Lin Xing Qi Hongde Li Jingjing Yang |
| author_sort | Siying Wang |
| collection | DOAJ |
| description | In recent years, the major geological hazard of landslides has greatly impact normal human life. Deep belief networks (DBN) is a commonly used deep learning model, and the DBN hyperparameter determination problem is the key to its application. To improve the accuracy of regional landslide susceptibility prediction, this paper introduces the particle swarm algorithm (PSO) to determine the hyperparameters of the DBN; this is applied to regional landslide susceptibility prediction. Firstly, PSO is used to optimize the hyperparameters of the DBN and obtain a set of hyperparameters with the optimal fitness function. A landslide susceptibility prediction model based on PSO-DBN is then constructed and the K-fold cross-validation method is used to determine its accuracy. The model is applied to landslide susceptibility prediction in the most impacted area of the Wenchuan earthquake to analyze the model’s accuracy. Finally, model susceptibility analysis is performed. The research results show that the final optimal model accuracy of the PSO-DBN model is 95.52%, which is approximately 28.31% and 15.35% higher than that of the logistic regression (LR) model and the common DBN model, respectively. The Kappa coefficient is 0.883, which is higher than that of the LR model. Compared with the LR model and the common DBN model, Kappa coefficient is improved by approximately 0.542 and 0.269 respectively; the area under the curve (AUC) is 0.951, which is improved by approximately 0.201 and 0.080 compared to the LR model and the common DBN model. The susceptibility of the model to the inertia factor is low, the average change in model accuracy (when the inertia factor changes by 0.1) is approximately 0.1%, and the overall stability of the model is high. The landslide susceptibility level is very high. The area includes 219 landslide points, which account for 39.2% of total landslide points. In the area with a high level of landslide susceptibility are 191 landslide points, accounting for 34.2% of total landslide points. Together, the two contain approximately 73.4% of the landslide points. This indicates that the model prediction results agree well with the spatial distribution characteristics of the landslide. |
| first_indexed | 2024-04-14T02:53:41Z |
| format | Article |
| id | doaj.art-6f9b471eddc744cdbf2505c3f9d287b1 |
| institution | Directory Open Access Journal |
| issn | 2296-665X |
| language | English |
| last_indexed | 2024-04-14T02:53:41Z |
| publishDate | 2022-08-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Environmental Science |
| spelling | doaj.art-6f9b471eddc744cdbf2505c3f9d287b12022-12-22T02:16:12ZengFrontiers Media S.A.Frontiers in Environmental Science2296-665X2022-08-011010.3389/fenvs.2022.912523912523Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken areaSiying Wang0Xiaokun Lin1Xiaokun Lin2Xiaokun Lin3Xing Qi4Hongde Li5Jingjing Yang6College of Management Science, Chengdu University of Technology, Chengdu, ChinaCollege of Management Science, Chengdu University of Technology, Chengdu, ChinaCollege of Mathematics and Physics, Chengdu University of Technology, Chengdu, ChinaGeomathematics Key Laboratory of Sichuan Province, Chengdu, ChinaCollege of Civil Engineering, Sichuan University of Science and Engineering, Zigong, ChinaCollege of Mathematics and Physics, Chengdu University of Technology, Chengdu, ChinaCollege of Mathematics and Physics, Chengdu University of Technology, Chengdu, ChinaIn recent years, the major geological hazard of landslides has greatly impact normal human life. Deep belief networks (DBN) is a commonly used deep learning model, and the DBN hyperparameter determination problem is the key to its application. To improve the accuracy of regional landslide susceptibility prediction, this paper introduces the particle swarm algorithm (PSO) to determine the hyperparameters of the DBN; this is applied to regional landslide susceptibility prediction. Firstly, PSO is used to optimize the hyperparameters of the DBN and obtain a set of hyperparameters with the optimal fitness function. A landslide susceptibility prediction model based on PSO-DBN is then constructed and the K-fold cross-validation method is used to determine its accuracy. The model is applied to landslide susceptibility prediction in the most impacted area of the Wenchuan earthquake to analyze the model’s accuracy. Finally, model susceptibility analysis is performed. The research results show that the final optimal model accuracy of the PSO-DBN model is 95.52%, which is approximately 28.31% and 15.35% higher than that of the logistic regression (LR) model and the common DBN model, respectively. The Kappa coefficient is 0.883, which is higher than that of the LR model. Compared with the LR model and the common DBN model, Kappa coefficient is improved by approximately 0.542 and 0.269 respectively; the area under the curve (AUC) is 0.951, which is improved by approximately 0.201 and 0.080 compared to the LR model and the common DBN model. The susceptibility of the model to the inertia factor is low, the average change in model accuracy (when the inertia factor changes by 0.1) is approximately 0.1%, and the overall stability of the model is high. The landslide susceptibility level is very high. The area includes 219 landslide points, which account for 39.2% of total landslide points. In the area with a high level of landslide susceptibility are 191 landslide points, accounting for 34.2% of total landslide points. Together, the two contain approximately 73.4% of the landslide points. This indicates that the model prediction results agree well with the spatial distribution characteristics of the landslide.https://www.frontiersin.org/articles/10.3389/fenvs.2022.912523/fullAUCdeep belief network (DBN)PSO—particle swarm optimizationK-fold cross-validation methodlandslide susceptibility |
| spellingShingle | Siying Wang Xiaokun Lin Xiaokun Lin Xiaokun Lin Xing Qi Hongde Li Jingjing Yang Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area Frontiers in Environmental Science AUC deep belief network (DBN) PSO—particle swarm optimization K-fold cross-validation method landslide susceptibility |
| title | Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area |
| title_full | Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area |
| title_fullStr | Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area |
| title_full_unstemmed | Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area |
| title_short | Landslide susceptibility analysis based on a PSO-DBN prediction model in an earthquake-stricken area |
| title_sort | landslide susceptibility analysis based on a pso dbn prediction model in an earthquake stricken area |
| topic | AUC deep belief network (DBN) PSO—particle swarm optimization K-fold cross-validation method landslide susceptibility |
| url | https://www.frontiersin.org/articles/10.3389/fenvs.2022.912523/full |
| work_keys_str_mv | AT siyingwang landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea AT xiaokunlin landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea AT xiaokunlin landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea AT xiaokunlin landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea AT xingqi landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea AT hongdeli landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea AT jingjingyang landslidesusceptibilityanalysisbasedonapsodbnpredictionmodelinanearthquakestrickenarea |