Analysis of the composition of ancient glass objects based on cluster analysis and random forest methods

Zhiying Chen; Dongchen Shang

doi:10.54097/hset.v40i.6506

Authors

Zhiying Chen
Dongchen Shang

DOI:

https://doi.org/10.54097/hset.v40i.6506

Keywords:

Cluster analysis, Random Forest, Glass classification study.

Abstract

In ancient times fluxes were often added to the production of glass to lower the melting point of pure quartz sand. During the smelting process, the addition of different products resulted in significant changes in the internal composition of the glass products. This paper examines a sample of ancient glassworks, based on the data from several sources. The data are divided into 'weathered' and 'unweathered', and cluster analysis reveals significant differences between the categories, and random forest analysis is used to determine the non-linear relationship between the variables, resulting in a chemical composition correlation of the relationships were more variable for potassium chloride and lead chloride and smoother for the rest. This paper innovatively adopts a combination of K-means cluster analysis and random forest to evaluate the composition and identification problems of ancient glass objects. The model in this paper can also be extended to other studies related to the degree of weathering in ancient wooden shipwrecks, metals, ceramics, and other aspects.

Downloads

Download data is not yet available.

References

Liang J, Chen JH, Zhang XUEQIN, Zhou YUE, Lin JIAJUN. Anomaly detection based on unique thermal coding and convolutional neural network[J]. Journal of Tsinghua University (Natural Science Edition), 2019, 59 (07).

WANG Quan, CHENG Xiaofang, FU Teran, LU Shaosong. Definition of the colour gamut of continuous radioluminescence in the visible wavelength band [J]. Science Bulletin, 2002, (13): 972 - 977.

Zhou, Nana, Rao, Zhijian. Analysis of factors influencing grain yield in Yunnan Province based on grey correlation analysis [J]. Agriculture and Technology, 2022, 42 (15): 164 - 167.

Zhang Tie, Chen Jun, Xue Chunzhu, Mu Cunfu. Value analysis of a random forest algorithm-based prediction model for acute kidney injury in postoperative cardiothoracic patients [J]. Journal of Cardiology, 2023, (01): 67 - 71.

He Wenliang, Fu Lianlian, Liao Jingping. Research on pig price forecasting and regulation mechanism based on random forest model [J]. Price Monthly: 1 - 10.

Li Y R. Application of random forests in agriculture [J]. Southern Agricultural Machinery, 2022, 53 (22): 63 - 65+87.

Xia Shuyuan, Dong Yongfeng, Wang Liqin. Research on XGBoost blast block prediction based on feature engineering [J]. Blasting: 1 - 9.

Zhou Yisong, Zhao Chuanping, Huang Yaoming, Zhu Li, Cheng Ming. Research on the prediction of compressive strength of concrete based on machine learning technology [J]. Journal of Anyang Engineering College, 2022, 21 (06): 91 - 95.

XU Qingyu, YU Jing, ZHU Dawei, ZHENG Xiaolong, MENG Lingqi, ZHU Zhiwei, SHAO Yafang. Study on the evaluation of nutritional quality of different rice varieties based on principal component analysis and cluster analysis [J]. China Rice,2022, 28 (06): 1 - 8.

Liu Shuna. Research on glassware of nomadic peoples in ancient China's north [D]. Inner Mongolia Normal University, 2022.