基于回归方法的玻璃文物成分鉴定
Compositional Identification of Glass Artifacts Based on Regression and Machine Learning
郑铛 1陈煜 1林建豪 1夏志乐1
作者信息
- 1. 台州学院 电子与信息工程学院,浙江 台州 318000
- 折叠
摘要
基于全国大学生数学建模竞赛赛题中的玻璃文物数据,采用Lasso回归、向后逐步回归、普通最小二乘法(OLS)的稳健标准误差回归等回归模型,构建高钾玻璃与铅钡玻璃的线性回归方程,预测表面风化文物在风化之前的化学成分质量分数,以鉴定玻璃文物的成分.结果表明,该模型的预测结果较合理.
Abstract
Based on the data of glass artifacts of the CUMCM,regression models such as Lasso regression,backward stepwise regression and robust standard error regression by ordinary least squares(OLS)were used to give linear regres-sion equations for high-potassium and lead-barium glasses,and to identify the chemical composition content of surface weathered artifacts prior to weathering.
关键词
玻璃文物/风化/多元回归Key words
glass artifacts/weathering/multiple regression引用本文复制引用
出版年
2024
NETL
NSTL
维普
万方数据