摘要：在当今互联网高速发展的局势下，随之兴起的是各式各样的网络软件产品，众包已然变成了一种新型的电子商务模式，并拥有了一定的市场份额。本论文是对数模竞赛题——APP拍照定价系统，这种自助式众包平台任务定价系统的进一步研究。主要通过历史任务记录、会员信息等数据来寻找任务是否完成的关键因素，以及分析确定最终的定价模型。研究过程中分别对会员坐标、历史任务坐标做了聚类分析，得出四个聚类中心，分别位于深圳、佛山、东莞、广州四个城市。再通过网络采集这四个城市本身的经济、人口发展水平，决定系统在不同城市的基础定价。再对已有数据做数据规约，获取想要的、直观的数据，并对这些数据做多元线性回归，得到回归系数，确立最终定价回归方程，最终得出结论。问题是否完成主要与任务定价、任务附件会员密度、任务离群度密切相关。论文完成后发现对经纬度的距离计算没有采取专门的计算公式，可能会导致最终数据会有一定的偏差。

关键词： 定价模型；聚类分析；网络采集；数据规约；多元线性回归

Analysis of APP Camera Pricing System

Abstract: In the current situation of the rapid development of the Internet, various kinds of network software products are emerging. Crowdsourcing has become a new type of e-commerce model and has a certain market share. This dissertation is a further study of the digital model contest title—APP camera pricing system, which is a self-service crowdsourcing platform task pricing system. Mainly through historical mission records, membership information and other data to find the key factors of the completion of the task, and analysis to determine the final pricing model. In the research process, cluster analysis was performed on the coordinates of the members and the coordinates of historical missions, and four clustering centers were found, which were located in Shenzhen, Foshan, Dongguan and Guangzhou. Through the network, the economic and demographic development levels of the four cities themselves are collected, and the basic pricing of the system in different cities is determined. Data stipulations were made on existing data and desired and intuitive data were obtained. Multiple linear regressions were performed on these data to obtain regression coefficients, and final pricing regression equations were established. Finally, conclusions were reached. Whether the problem is completed is mainly related to the pricing of the task, the density of the task attachment membership, and the degree of task outlier. After the paper is completed, it is found that there is no specific calculation formula for distance calculation of latitude and longitude, which may cause certain deviation of the final data.

Keywords: Pricing model; Cluster analysis; Network acquisition; Data specification; Multiple linear regression

目录

绪论 1

一．数据处理 3

1.1数据集 3

1.2数据清洗 3

1.3数据规约 3

1.3.1任务附件会员密度 3

1.3.2任务附近会员平均距离 3

1.3.3任务距离聚类中心点距离 4

二.需求分析 4

2.1任务未完成的原因 4

2.2 完整的定价模型 4

三.模型建立 4

3.1 假设 4

3.2定价系统分析 4

3.2.1流程图 4

3.2.2 任务散点图 5

3.2.3任务完成率 5

3.3模型建立 6

3.4定价规律的分析 6

3.4.1 任务附近会员的密度分布 6

3.4.2任务附近会员平均距离 6

3.4.3距离聚类中心的距离 6

四.回归分析 7

4.1多元线性回归 7

4.2异方差检验 8

4.3定价规律总结 8

五.实际因素 8

5.1经济状况 8

5.2任务吸引度 9

5.3任务限额 9

5.4开始预计时间 9

5.5会员信誉 9

六.定价研究 9

6.1基础定价 9

6.2标价基础值 10