摘要
为了准确预测儿童服装流行趋势,针对传统预测方式中存在的主观性问题,提出了基于BP神经网络的童装流行元素预测模型。该预测模型以国内淘宝、天猫、京东、苏宁易购等主要电商平台在2000—2020年间的历史销售数据作为反映流行程度的指标,以童装流行色作为预测案例,采用虚拟变量的方法,对童装的造型、款式、材料、色彩、图案、结构、工艺、搭配和风格九大流行元素进行量化,并使用MatLab平台编写程序建立预测模型,对样本数据网络进行训练,调整隐含层节点数,以BP神经网络模型模拟预测结果,对2021—2022年秋冬的童装流行色三要素进行预测并输出。得出的预测结果与市场流行趋势一致。
In order to accurately predict the fashion trend of children′s wear,a prediction model based on BP neural network for children′s clothing fashion elements was proposed to address the subjectivity problem in the traditional prediction methods.The historical sales data of major domestic e-commerce platforms such as Taobao,Tmall,Jingdong and Suning Tesco during the period 2000—2020 was used as indicators to reflect the degree of popularity in the prediction model,children′s fashion colors were taken as prediction cases,and a dummy variable approach was adopted and quantified to the nine children′s fashion elements of children′s clothing:shape,style,material,color,pattern,structure,process,matching and style.The prediction model was built using the MatLab platform,the sample data network was trained,the number of nodes in the implicit layer was adjusted,and the prediction results were simulated with a BP neural network model to predict and output the three elements of children′s fashion colors for autumn and winter 2021—2022.The prediction results obtained are consistent with the market trends.
作者
刘妍兵
刘伦伦
唐颖
LIU Yanbing;LIU Lunlun;TANG Ying(School of Design,Jiangnan University,Wuxi,Jiangsu 214122,China;School of Design,Polytechnic Di Milano,Milan 20158,Italy)
出处
《毛纺科技》
CAS
北大核心
2022年第2期109-115,共7页
Wool Textile Journal
基金
教育部人文社会科学研究青年基金项目(19YJC760096)
国家重点研发计划重点专项(2019YFB1405700)
国家留学基金资助项目(201806795029)。
关键词
BP神经网络
童装
流行元素
流行色
预测分析
BP neural network
children′s wear
fashion elements
color trends
prediction
