GIS-based non-grain cultivated land susceptibility prediction using data mining methods
Abstract The purpose of the present study is to predict and draw up non-grain cultivated land (NCL) susceptibility map based on optimized Extreme Gradient Boosting (XGBoost) model using the Particle Swarm Optimization (PSO) metaheuristic algorithm. In order to, a total of 184 NCL areas were identifi...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-55002-y |
| _version_ | 1797274597508251648 |
|---|---|
| author | Qili Hao Tingyu Zhang Xiaohui Cheng Peng He Xiankui Zhu Yao Chen |
| author_facet | Qili Hao Tingyu Zhang Xiaohui Cheng Peng He Xiankui Zhu Yao Chen |
| author_sort | Qili Hao |
| collection | DOAJ |
| description | Abstract The purpose of the present study is to predict and draw up non-grain cultivated land (NCL) susceptibility map based on optimized Extreme Gradient Boosting (XGBoost) model using the Particle Swarm Optimization (PSO) metaheuristic algorithm. In order to, a total of 184 NCL areas were identified based on historical records, and a total of 16 NCL susceptibility conditioning factors (NCLSCFs) were considered, based on both a systematic literature survey and local environmental conditions. The results showed that the XGBoost model optimized by PSO performed well in comparison to other machine learning algorithms; the values of sensitivity, specificity, PPV, NPV, and AUC are 0.93, 0.89, 0.88, 0.93, and 0.96, respectively. Slope, rainfall, fault density, distance from fault and drainage density are most important variables. According to the results of this study, the use of meta-innovative algorithms such as PSO can greatly enhance the ability of machine learning models. |
| first_indexed | 2024-03-07T15:00:38Z |
| format | Article |
| id | doaj.art-a4a20c8dcae843ef96da6bf47301ce0a |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-07T15:00:38Z |
| publishDate | 2024-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-a4a20c8dcae843ef96da6bf47301ce0a2024-03-05T19:11:15ZengNature PortfolioScientific Reports2045-23222024-02-0114111710.1038/s41598-024-55002-yGIS-based non-grain cultivated land susceptibility prediction using data mining methodsQili Hao0Tingyu Zhang1Xiaohui Cheng2Peng He3Xiankui Zhu4Yao Chen5Shangluo Branch, Shaanxi Provincial Land Engineering Construction GroupShangluo Branch, Shaanxi Provincial Land Engineering Construction GroupShangluo Branch, Shaanxi Provincial Land Engineering Construction GroupShangluo Tea Research InstituteShangluo Tea Research InstituteShangnan County Tea Industry Development CenterAbstract The purpose of the present study is to predict and draw up non-grain cultivated land (NCL) susceptibility map based on optimized Extreme Gradient Boosting (XGBoost) model using the Particle Swarm Optimization (PSO) metaheuristic algorithm. In order to, a total of 184 NCL areas were identified based on historical records, and a total of 16 NCL susceptibility conditioning factors (NCLSCFs) were considered, based on both a systematic literature survey and local environmental conditions. The results showed that the XGBoost model optimized by PSO performed well in comparison to other machine learning algorithms; the values of sensitivity, specificity, PPV, NPV, and AUC are 0.93, 0.89, 0.88, 0.93, and 0.96, respectively. Slope, rainfall, fault density, distance from fault and drainage density are most important variables. According to the results of this study, the use of meta-innovative algorithms such as PSO can greatly enhance the ability of machine learning models.https://doi.org/10.1038/s41598-024-55002-yMetaheuristic algorithmsParticle swarm optimizationOptimized extreme gradient boostingEnvironmental management |
| spellingShingle | Qili Hao Tingyu Zhang Xiaohui Cheng Peng He Xiankui Zhu Yao Chen GIS-based non-grain cultivated land susceptibility prediction using data mining methods Scientific Reports Metaheuristic algorithms Particle swarm optimization Optimized extreme gradient boosting Environmental management |
| title | GIS-based non-grain cultivated land susceptibility prediction using data mining methods |
| title_full | GIS-based non-grain cultivated land susceptibility prediction using data mining methods |
| title_fullStr | GIS-based non-grain cultivated land susceptibility prediction using data mining methods |
| title_full_unstemmed | GIS-based non-grain cultivated land susceptibility prediction using data mining methods |
| title_short | GIS-based non-grain cultivated land susceptibility prediction using data mining methods |
| title_sort | gis based non grain cultivated land susceptibility prediction using data mining methods |
| topic | Metaheuristic algorithms Particle swarm optimization Optimized extreme gradient boosting Environmental management |
| url | https://doi.org/10.1038/s41598-024-55002-y |
| work_keys_str_mv | AT qilihao gisbasednongraincultivatedlandsusceptibilitypredictionusingdataminingmethods AT tingyuzhang gisbasednongraincultivatedlandsusceptibilitypredictionusingdataminingmethods AT xiaohuicheng gisbasednongraincultivatedlandsusceptibilitypredictionusingdataminingmethods AT penghe gisbasednongraincultivatedlandsusceptibilitypredictionusingdataminingmethods AT xiankuizhu gisbasednongraincultivatedlandsusceptibilitypredictionusingdataminingmethods AT yaochen gisbasednongraincultivatedlandsusceptibilitypredictionusingdataminingmethods |