...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>International Journal of Environmental Research and Public Health >Genetic k-Means Clustering Approach for Mapping Human Vulnerability to Chemical Hazards in the Industrialized City: A Case Study of Shanghai, China
【24h】

Genetic k-Means Clustering Approach for Mapping Human Vulnerability to Chemical Hazards in the Industrialized City: A Case Study of Shanghai, China

机译：遗传k-均值聚类方法在工业化城市中绘制人类易感性化学危险图：以中国上海为例

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Reducing human vulnerability to chemical hazards in the industrialized city is a matter of great urgency. Vulnerability mapping is an alternative approach for providing vulnerability-reducing interventions in a region. This study presents a method for mapping human vulnerability to chemical hazards by using clustering analysis for effective vulnerability reduction. Taking the city of Shanghai as the study area, we measure human exposure to chemical hazards by using the proximity model with additionally considering the toxicity of hazardous substances, and capture the sensitivity and coping capacity with corresponding indicators. We perform an improved k-means clustering approach on the basis of genetic algorithm by using a 500 m × 500 m geographical grid as basic spatial unit. The sum of squared errors and silhouette coefficient are combined to measure the quality of clustering and to determine the optimal clustering number. Clustering result reveals a set of six typical human vulnerability patterns that show distinct vulnerability dimension combinations. The vulnerability mapping of the study area reflects cluster-specific vulnerability characteristics and their spatial distribution. Finally, we suggest specific points that can provide new insights in rationally allocating the limited funds for the vulnerability reduction of each cluster.
机译：在工业化城市中降低人们对化学危害的脆弱性是当务之急。漏洞映射是在区域中提供减少漏洞干预措施的替代方法。这项研究提出了一种通过使用聚类分析来有效减少脆弱性的方法将人类脆弱性映射到化学危害的方法。我们以上海市为研究区域，通过邻近模型测量人类暴露于化学危险中的程度，另外还考虑了有害物质的毒性，并通过相应的指标捕获了敏感性和应对能力。我们使用500 m×500 m地理网格作为基本空间单位，在遗传算法的基础上执行了改进的k-means聚类方法。平方误差的总和与轮廓系数相结合，以测量聚类的质量并确定最佳聚类数。聚类结果揭示了一组六个典型的人类漏洞模式，这些模式显示了不同的漏洞维度组合。研究区域的漏洞映射反映了特定于群集的漏洞特征及其空间分布。最后，我们提出一些具体观点，这些观点可以为合理分配有限的资金用于减少每个群集的漏洞提供新的见解。

著录项

相似文献

  • 外文文献
  • 中文文献
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号