基于多元线性回归对众包任务价格的设定  

Analysis and evaluation on the price of crowdsourcing assignments

作　　者：李嵩松[1] 赖博轩 陈驰杰 孙顶 LI Song-song;LAI Bo-xuan;CHEN Chi-jie;SUN Ding(School of Management,Harbin Institute of Technology,Harbin 150001,China;School of Electrical Engineering and Automation,Harbin Institute of Technology,Harbin 150001,China)

机构地区：[1]哈尔滨工业大学管理学院,黑龙江哈尔滨150001 [2]哈尔滨工业大学电气工程及自动化学院,黑龙江哈尔滨150001

出　　处：《齐齐哈尔大学学报（自然科学版）》2019年第1期78-84,共7页Journal of Qiqihar University(Natural Science Edition)

基　　金：国家自然科学基金(71773024)

摘　　要：在空间众包研究中,针对网络众包任务的价问题,综合分析众包任务发布地的会员密度、商业发展等数据,通过建立多元回归线性分析模型,得出对网络众包任务的价策略。通过聚类分析对任务的发布做优化,对一些距离过近的任务进行打包。最后利用神经网络算法由已完成的任务数据,预测待完成的任务是否成功。通过对预测数据进行仿真,得出任务密度、会员密度、会员信誉度和价对任务完成度与任务价成本的影响。In this paper,based on the data of the member density of crowdsourcing assignments and commercial statistics,a multifactor linear regression model is established and so is the pricing strategy.Moreover,the optimization of the assignments is done according to clustering analysis.At last,a neural network algorithm for evaluating the price whether they can be accepted by the customer is given as well.Based on those models,the solution of the simulation sees the impact which the density of assignments and members,credit and price bring to the completion ratio and the cost of the platform.

关 键 词：网络众包 多元线性回归分析 K均值聚类分析 神经网络 

分 类 号：O29[理学—应用数学]