基于优化GridSearch-SVM算法的服装版型分类预测研究

Research on Classification and Prediction of Garment Pattern Based on GridSearch-SVM Algorithm

服装定制推荐在服装定制过程中有着重要意义,要实现定制推荐的过程,首先根据其量体数据预测服装版型。根据某服装企业的量体数据与版型数据之间的组合关系,采用SVM分类算法,训练分类模型,根据量体数据预测出服装版型,并提出一种二度网格搜索算法(GridSearch),进行支持向量机的(C,g)参数寻优,实现最优的分类效果,并缩短运算时间。

Clothing customization recommendation played an important role in the process of clothing customization.In order to realize the process of customization recommendation,the clothing pattern was predicted according to body measure data.According to the combination relationship between the body-measure data and the pattern data of a clothing enterprise,SVM classification algorithm was used to train the classification model,and the clothing pattern was predicted according to the body-measure data.A two-dimensional grid search was proposed.The(C,g) parameters of support vector machine was optimized,to achieve the optimal classification effect and shorten the operation time.