摘要
基于人体皮肤状态提出相适应的护肤方案,是皮肤领域亟待解决的问题。采用专家评分法、标准离差法及熵权法融合主客观信息优化面部皮肤各部位权重,较全面地获取皮肤属性值;将被测试对象的护肤偏好与专家的方案—属性特征评价信息相结合,构建目标—方案—属性层次模型,实现个性偏好与理性推荐下的多属性权值调整;依据实测数据,采用多属性权值对TOPSIS法进行改进,通过计算各方案与正负理想方案之间的加权距离实现护肤方案的优劣排序。最后,通过一个应用实例以及与现有的两种决策方法的对比分析验证了方法的可行性与有效性。
Proposing a suitable skin care regimen based on the skin condition of human body is an urgent problem to be disposed of.This paper adopted the expert grading method,standard deviation method and entropy weight method,aimed at optimizing the position weight value of the facial skin,and acquired skin attribute values more comprehensively.It established goal-project-attribute hierarchical model,which combining the preference of tested sample and project-attribute characteristic by expert,to realize multiple attribute weights adjustment under individual preferences and rational recommendation.By analyzing measured data,it calculated the weighted distance between the schemes and the positive and negative ideal schemes utilizing optimized TOPSIS method dependent upon multi attribute weights,to realize the ranking of skin care programs.Finally,it verifies the feasibility and effectiveness of proposed method through an application example and comparison analysis with two exis-ting methods.
作者
张欣
王小艺
张慧妍
许继平
Zhang Xin;Wang Xiaoyi;Zhang Huiyan;Xu Jiping(Key Laboratory of Cosmetic of China National Light Industry,School of Computer&Information Engineering,Beijing Technology&Business University,Beijing 100048,China)
出处
《计算机应用研究》
CSCD
北大核心
2018年第11期3302-3306,共5页
Application Research of Computers
基金
北京工商大学促进人才培养综合改革项目(19005428069/007)
北京市属高校创新能力提升计划项目(PXM2014_014213_000033)
北京市教委科技重点项目(KZ201510011011)
关键词
最小二乘法
偏好信息
层次模型
皮肤
least square method
preference information
hierarchical model
skin
