﻿ 电商平台线上女装销售量的建模与预测 Modeling and Forecasting of Online Women’s Sales Volume in ECommerce Platform

Statistics and Application
Vol. 08  No. 02 ( 2019 ), Article ID: 29922 , 16 pages
10.12677/SA.2019.82045

Modeling and Forecasting of Online Women’s Sales Volume in E-Commerce Platform

Rongxiang Zhao, Yichao Yue, Siqi Liu, Lintao Zheng, Jin Yang

Dalian Minzu University, Dalian Liaoning

Received: Apr. 4th, 2019; accepted: Apr. 19th, 2019; published: Apr. 26th, 2019

ABSTRACT

This paper thinks that the sales data of Rakumachi, Only, and Taiping Bird in 2014-2017 can be regarded as a time series. First, the type of time series is determined by time series timing diagram. Then, the data is preprocessed according to different types. Since the sequence has a cluster effect, the heteroscedasticity is eliminated by taking the logarithm. Then, a preliminary mathematical model is proposed in turn, and linear fitting and residual fitting are performed to build a residual autoregressive model. By establishing the most appropriate time series model for the model to continuously identify, fit, test and optimize, the predicted value of the logarithm of the sales volume is obtained, and the sales volume of the three brands is predicted in the next six months.

Keywords:Electronic Business Platform, Time Series Model, Timing Chart Test, Linear Fit, Forecast

1. 前言

2. 问题分析

3. 名词解释

1) 时序图：亦称为序列图或循序图，是一种UML行为图。它通过描述对象之间发送消息的时间顺序显示多个对象之间的动态协作。

2) 时间序列模型：时间序列模型，是在生产和科学研究中，对某一个或一组变量 ${x}_{\left(t\right)}$ 进行观察测量，将在一系列时刻 ${t}_{1},{t}_{2},\cdots ,{t}_{n}$ (t为自变量)按照时间次序排列，并用于解释变量和相互关系的数学表达式。所得到的离散数字组成序列集合 $x\left({t}_{1}\right),x\left({t}_{2}\right),\cdots ,x\left({t}_{n}\right)$ ，我们称之为时间序列，这种有时间意义的序列也称为动态数据。根据对系统进行观测得到的时间序列数据，用曲线拟合方法对系统进行客观的描述。

4. 时间序列模型的建立

4.1. 时序图检验

Figure 1. Only timing chart

Figure 2. Taiping Bird timing chart

Figure 3. Rakumachi timing chart

4.2. 差分

Figure 4. Only timing chart

Figure 5. Taiping Bird timing chart

Figure 6. Rakumachi timing chart

4.3. 消除异方差性

4.4. 拟合时间序列模型

4.4.1. 线性拟合及残差检验

Figure 7. Only timing chart

Figure 8. Taiping Bird timing chart

Figure 9. Rakumachi timing chart

Figure 10. Only timing chart

${x}_{t}=10.816522+0.039712t+{\epsilon }_{t}$

Figure 11. Taiping Bird timing chart

${x}_{t}=11.256435+0.02883t+{\epsilon }_{t}$

Figure 12. Rakumachi timing chart

${x}_{t}=10.084547+0.041794t+{\epsilon }_{t}$

4.4.2. 残差拟合

Figure 13. Only residual timing chart

Figure 14. Autocorrelation chart

Figure 15. Partial autocorrelation chart

$\left\{\begin{array}{c}{x}_{t}=10.816522+0.039712t+{\epsilon }_{t}\\ {\epsilon }_{t}=-0.2631{\epsilon }_{t-1}+0.8118{e}_{t-1}+{e}_{t},{e}_{t}~N\left(0,0.1828\right)\end{array}$

$\left\{\begin{array}{c}{x}_{t}=11.256435+0.02883t+{\epsilon }_{t}\\ {\epsilon }_{t}=0.4492{\epsilon }_{t-1}+{e}_{t},{e}_{t}~N\left(0,08884\right)\end{array}$

Figure 16. Taiping Bird residual timing chart

Figure 17. Autocorrelation chart

Figure 18. Partial autocorrelation chart

Figure 19. Rakumachi residual timing chart

Figure 20. Autocorrelation Chart

Figure 21. Partial autocorrelation chart

$\left\{\begin{array}{c}{x}_{t}=10.084547+0.041794t+{\epsilon }_{t}\\ {\epsilon }_{t}=-0.5638{\epsilon }_{t-1}+0.9427{e}_{t-1}+{e}_{t},{e}_{t}~N\left(0,0.1155\right)\end{array}$

5. 预测

#only

Figure 22. Tendency chart

#太平鸟Taiping Bird

6. 总结

Figure 23. Tendency chart

Figure 24. Tendency chart

Modeling and Forecasting of Online Women’s Sales Volume in ECommerce Platform[J]. 统计学与应用, 2019, 08(02): 404-419. https://doi.org/10.12677/SA.2019.82045

1. 1. 蔡建梅, 李易儒. 基于淘宝网络销售平台的原创女装品牌发展策略研究[J]. 浙江理工大学学报, 2012, 29(4): 526-531.

2. 2. 万艳敏, 黄珊, 张斯雯. 女装品牌直营终端销售业务流程指标体系研究[J]. 北京服装学院学报, 2013, 33(2): 12-17.

3. 3. 文春英, 张淑梅. 网络购物说服策略的运用-基于淘宝网女装销售文本的内容分析[J]. 新媒体, 2013(5): 77-80.

4. 4. 李家华, 王睿, 陈佩芳. 浅谈电子商务模式下女装服饰销售策略分析[J]. 西部皮革, 2016, 38(12): 71-71.

5. 5. 王燕. 时间序列分析: 基于R [M]. 北京: 中国人民大学出版社, 2015: 94, 162.

6. 6. 中华全国商业信息中心. 女装上位[J]. 商业消费趋势, 2014(7): 92-93.

7. 7. 艾维娜, 杨坚争. B2B电子商务平台发展及存在的问题[J]. 电子商务, 2018(1): 9-10.

8. 8. 女装商战跳过春装买夏装[J]. 纺织服装周刊, 2015(15): 77.