基于区块链的毕业设计KMeans Clustering using Covid and Socio-Economic Data – 利用Covid和社会经济数据进行KMeans聚类

本文提供基于区块链的毕业设计国外最新区块链项目源码下载,包括solidity,eth,fabric等blockchain区块链,基于区块链的毕业设计KMeans Clustering using Covid and Socio-Economic Data – 利用Covid和社会经济数据进行KMeans聚类 是一篇很好的国外资料

KMeans Clustering using Covid and Socio-Economic Data

A look at how clustering algorithms like KMeans can help us better understand Covid-19’s affects on different US counties

Matt Paterson, hello@hireMattPaterson.com
Machine Learning Engineer
Cloud Brigade, Santa Cruz, CA

Web API

https://public.tableau.com/views/Covid_Clustering_SocioEconomic_Racial_Data/Overview?:language=en&:display_count=y&publish=yes&:origin=viz_share_link

See an easy to use Tableau Dashboard here

The Data Science Question

Can we use Unsupervised Machine Learning models to learn new things about Covid-19 and how it is affecting different counties in different parts of the country?

We set out to assemble as much data from each US county as possible as a way to see what each county might have in common and how that influences a county’s propensity for and the severity of Covid-19. Here, we use the recorded number of confirmed cases as well as deaths in each county, and added to that economic data and ethnic population data from the Beaureau of Economic Analysis and pumped all of this data in to a KMeans clustering algorithm.

Top-Level Results

The findings were interesting but really not anything new. To whit: Counties with higher African American populations are most likely to have higher deaths per capita as a result of Covid than counties with low African American populations. That was the single strongest feature of correlation in our data.

As you can see from the Tableau API, or from a run through the clustering modeling in the KMeans ipynb, the geographic areas that have been hardest hit by the pandemic include high-density city cores as well as very low-density rural areas. The one thing that all of these areas have in common is their non-white racial ratios, and specifically a high relative percentage of African American people versus the rest of the population.

This project was never meant to be political but only to look at the numbers and listen to what the data tell us. So far, this is what we see. This study does not suggest the causation of this correlation, only the existence of it. Whether the reasons have to do with underlying health issues (or the causes therein), the number of adults living in a single household, or the ability to work from home versus reporting to work every day with other people, this particular study was only concerned with finding patterns in the data and not finding the causation.


KMeans使用Covid和社会经济数据进行聚类

看看像KMeans这样的聚类算法如何帮助我们更好地理解Covid-19对美国不同县的影响

马特·帕特森,你好@hireMattPaterson.com
机器学习工程师
云旅,加利福尼亚州圣克鲁斯

Web API

https://public.tableau.com/views/Covid_Clustering_sociewed_种族_Data/Overview?:language=en&:display_count=y&publish=yes&origin=viz_share_link

在这里查看一个易于使用的Tableau仪表板

数据科学问题

我们是否可以使用无监督的机器学习模型来了解Covid-19的新知识,以及它如何影响全国不同地区的不同县?

我们开始收集尽可能多的来自美国每个县的数据,以了解每个县可能有哪些共同点,以及这如何影响一个县感染Covid-19的倾向和严重程度。在这里,我们使用每个县记录的确诊病例数和死亡人数,再加上来自经济分析的Beaureau的经济数据和民族人口数据,并将所有这些数据输入到KMeans聚类算法中。

顶层结果

这些发现很有趣,但并不是什么新鲜事。To whit:与非洲裔美国人人口较少的县相比,非洲裔美国人人口较多的县最有可能因Covid导致的人均死亡人数更高。这是我们数据中唯一最强大的相关性特征。

正如您从Tableau API或KMeans ipynb中的集群建模中所看到的,受流感大流行影响最严重的地理区域包括高密度的城市核心区以及非常低密度的农村地区。所有这些地区都有一个共同点,那就是他们的非白人种族比例,特别是非裔美国人与其他人口的相对比例较高。

这个项目从来就不是政治性的,只是看看数字,听听数据告诉我们什么。到目前为止,这就是我们所看到的。这项研究并没有指出这种相关性的原因,只是它的存在。无论原因是与潜在的健康问题(或其中的原因)、住在一个家庭的成年人的数量,还是在家工作与每天与其他人一起工作的能力有关,这项特别的研究只关注在数据中寻找模式,而不找出原因。

部分转自网络,侵权联系删除区块链源码网

www.interchains.cc

https://www.interchains.cc/17707.html

区块链毕设网(www.interchains.cc)全网最靠谱的原创区块链毕设代做网站 部分资料来自网络,侵权联系删除! 最全最大的区块链源码站 !
区块链知识分享网, 以太坊dapp资源网, 区块链教程, fabric教程下载, 区块链书籍下载, 区块链资料下载, 区块链视频教程下载, 区块链基础教程, 区块链入门教程, 区块链资源 » 基于区块链的毕业设计KMeans Clustering using Covid and Socio-Economic Data – 利用Covid和社会经济数据进行KMeans聚类

提供最优质的资源集合

立即查看 了解详情