聚类分析方法在我国空气污染
区域划分中的应用
安徽大学 笪婷婷、邹委员、武锦
摘 要
随着我国工业化进一步的进展，人们的生活也进一步的提高。伴随着经济的进展，环境也受到了一定的阻碍，国家也相应的提出了人与自然和谐相处的可持续进展战略。本文基于国家的政策，考虑工业化过程中都市空气的污染情况，提出相应的合理建议，从而使工业化过程中我们的环境也能受到更好的爱护！
聚类分析是目前最有前景的数据分析方法之一，它不仅能作为一个独立的工具来获得数据分布的情况，观看每一个簇的特点，还能集中地对某些特定的簇作进一步的分析。对空气污染区域划分的聚类分析，不仅能合理的分析我国各地区空气污染的情况，还能够对我们工业化进展的伟大蓝图提出我们瀚渺的建议，维护我们宽敞都市居民的切身利益！
本文首先对几种聚类方法进行了介绍和比较，然后在对我国空气污染现状分析中，运用了系统聚类分析方法。首先，我们采纳了组内连接聚类分析方法对我国的空气污染区域按照污染程度的不同进行了划分；其次，我们又采纳了中位数聚类分析法对我国的空气污染区域按照不同地区的废气处理情况进行了划分；最后我们又采纳了质心聚类分析法对各污染区域按年度的不同进行划分。我们用所选的三种数据进行聚类，产生的七个类是在整体上是一致的。这就表明，空气污染程度与废气处
1 / 51
理的力度是成正相关的。
为了讲明系统聚类分析方法在我国空气污染区域划分中的合理性，我们又采纳了k-means方法进行聚类，所得的聚类结果与运用系统聚类法的结果相似。从而进一步讲明分类的合理性。
依照聚类结果，我们提出了一些相关的防治空气污染的建议。相关部门应该按照污染地区的分类有针对性的制定相关策略，因地制宜，对污染程度相近的地区采纳合理的方式进行治理。
关键词：聚类方法；系统聚类法；中位数聚类分析；组内连接聚类分析；质 心聚类分析
2 / 51
The Application of Cluster Method in Air Pollution
Regional Division of Our Country
Abstract
With the further development of the industry in our country, people have undergone general improvement. However, the environment is accordingly suffering some attack from the progress of economy. The government has launched the strategy of sustainable development in order to build harmonious relationship between human and nature. Based on the policy of our government, taking the pollution in urban area into consideration, our article gives some corresponding advices to protect our environment in the industrialization.
Cluster method is one of the most promising methods in data analysis. Not only can it act as an independent tool to obtain the information of data distribution and observation of the characteristics of each cluster, but also do further analysis for some particular clusters. Using the cluster method in the analysis of division for polluted urban area, we can do reasonable analysis
3 / 51
and get acquaintance of condition of air pollution in different places. Our trivial recommendations will be come up for the blue sky of our industry development and safeguarding the vital interests of the city dwellers.
At the beginning of this paper, more than one kind of cluster method will be introduced and compared in this paper. Furthermore, systematic cluster methods will be applied in the division of the current air pollution circumstances. Firstly, we use team linked cluster method to divide regions according to the pollution degree. Secondly, we use median cluster method in accordance with the pollutant disposal of different areas. Lastly, we use centroid cluster method by judging the annual condition of air pollution. We cluster these three kinds of data, finding that seven categories we produce are accordant on the whole, which suggests that pollution degree is positively related to the strength we depose. To illustrate the rationality of systematic cluster method applied in air pollution regional division, we utilize k-means to cluster. To our happiness, the result we obtain is quite similar to that of systematic cluster method, which shows the rationality.
4 / 51
According to the clustering results, we put forward some relevant suggestions for the prevention and control of air pollution. Relevant departments should formulate relevant strategies based on the classification of pollution areas, and take suitable measures for local conditions. Also we should depose the pollution in similar levels in a reasonable way.
Keywords: : cluster system clustering method; median cluster analysis; connection cluster analysis within the group; a centroid cluster analysis
5 / 51
聚类分析方法在我国空气污染区域划分中的应用
