基于机器学习分类算法的玻璃文物鉴定  

Identification of Glass Cultural Relics Based on Machine Learning Classification Algorithm

作　　者：惠姣姣 曹红宝 HUI Jiaojiao;CAO Hongbao(School of Mathematics and Information Science,Baoji University of Arts and Sciences,Baoji 721013,China)

机构地区：[1]宝鸡文理学院数学与信息科学学院,陕西宝鸡721013

出　　处：《现代信息科技》2023年第13期101-104,共4页Modern Information Technology

基　　金：陕西省教育厅自然科学研究项目(202310592)。

摘　　要：近年来,玻璃文物受到的关注度越来越高。为了鉴定玻璃文物的类型、分析玻璃文物中的主要化学成分,该文基于高钾、铅钡玻璃的14种化学成分和表面风化情况,构建基于支持向量机(SVM)、人工神经网络(ANN)和Logistic回归(LR)的玻璃文物鉴定算法,对63种玻璃文物进行分类鉴定。结果显示:算法的预测精度均高于95%,可以将其应用到玻璃文物的实际鉴定中。In recent years,glass cultural relics have received more and more attention.In order to identify the types of glass cultural relics and analyze the main chemical components in glass cultural relics,based on 14 chemical components and surface weathering situation of high potassium and lead barium glass,this paper constructs a glass cultural relics identification algorithm based on Support Vector Machine(SVM),Artificial Neural Network(ANN)and Logistic Regression(LR),and classifies and identifies the 63 types of glass cultural relics.The results show that the prediction accuracy of the algorithm is higher than 95%,which can be applied to the actual identification of glass cultural relics.

关 键 词：玻璃文物鉴定 支持向量机 人工神经网络 LOGISTIC回归 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] G124[自动化与计算机技术—控制科学与工程]