基于Logistic回归的'拍照赚钱'APP定价方案设计

摘要

该文旨在对"拍照赚钱"APP考虑不同情况制定合理的定价方案.首先, 通过K-means聚类对定价的影响因素进行分析, 建立了以对数回归、多元线性回归为辅助解释的决策树定价模型, 从而得到原方案的定价规律, 并建立了任务完成情况的Logistic的模型, 结合统计分析与地理数据, 得到了影响任务完成的主要因素.随后, 在考虑总成本的前提下, 对定价模型进行优化, 对用户限额进行改进, 并基于Logistic模型以最大化任务完成率为目标对价格进行调整.最后, 设计了逐步聚类算法对任务进行打包, 进一步提高任务的完成率.%This paper aims to make a reasonable pricing plan for an"photo-taking"App considering different situations.Firstly, through the K-means clustering analysis of the influencing factors of pricing, a decision tree pricing model is established, which is explained by logarithmic regression and multiple linear regression as auxiliary.The pricing rule of the original plan is obtained, and the task completion situation is established.The Logistic model, combined with statistical analysis and geographic data, was the main factor that affected the task completion.Then, on the premise of considering the total cost, the pricing model is optimized, the user's quota is improved, and the price is adjusted based on the Logistic model to maximize the task completion rate.Finally, a step-by-step clustering algorithm was designed to package the task and further improve the completion rate of the task.