摘要
针对玻璃制品表面风化与各属性的相关性,对其进行化学成分含量预测与分类研究。以决策树、均值法为模型,根据数据计算纹饰、类型、颜色、表面是否风化的斯皮尔曼相关系数,得出类型与表面风化、纹饰、颜色之间有较弱相关性。利用卡方检验探究变量之间的差异性建立决策树模型,得出高钾玻璃、铅钡玻璃的分类规律,以分析决策树子节点。此决策树模型分类具有合理性,可使文物研究更加科学有效,促进玻璃制品的生产制造。
In view of the correlation between surface weathering and properties of glass products,the chemical composition content prediction and classification are studied.Through using decision tree and mean value method as models,Spearman correlation coefficients of pattern,type,color and surface weathering are calculated according to the data.It is concluded that there is weak correlation between type,and surface weathering,pattern and color.Chi-square test is used to explore the differences between variables to establish a decision tree model,and the classification rules of high-potassium glass and lead-barium glass are obtained to analyze the decision tree child nodes.The classification of decision tree model is reasonable,which can make the research of cultural relics more scientific and effective,and promote the production of glass products.
作者
翟司浔
Zhai Sixun(School of Information Technology,Hebei University of Economics and Business,Shijiazhuang 050000,China)
出处
《黑龙江科学》
2023年第8期47-49,共3页
Heilongjiang Science
关键词
决策树
均值法
卡方检验
相关系数
文物研究
玻璃制品
Decision tree
Mean value method
Chi-square test
Correlation coefficient
Cultural relic
Glass
