Research on Composition Analysis of Glass Products Based on Clustering and Random Forest Model

Jiarui Li; Jinglin Qi; Mei Yue

doi:10.54097/hset.v33i.5287

Authors

Jiarui Li
Jinglin Qi
Mei Yue

DOI:

https://doi.org/10.54097/hset.v33i.5287

Keywords:

Clustering algorithm, SOM, random forest model, machine learning.

Abstract

Glass cultural relics are important material and cultural heritage of human beings and play an important role in the history of human civilization. This study aims to investigate the relationship between the chemical composition of glass artefacts and the glass type of the artefacts. The chemical composition of glass was analyzed and identified by using K-means clustering, multiple correspondence analysis, SOM model and random forest algorithm. In this paper, based on the given data, the physical properties and chemical composition of the high-potassium glass relics and the lead-barium glass relics were studied respectively, and the classification basis of the two was obtained. It is found that the content of silica and lead oxide can be used as an important basis for distinguishing the two types of glass cultural relics and based on this, a prediction model for further classification of cultural relics and inference of glass cultural relics is designed. This study answered the relationship between the chemical composition of glass relics and the glass type of relics, and further research is needed to obtain more relevant data to train the model to improve its applicability.

Downloads

Download data is not yet available.

References

Yang Xin. Glass defect detection and classification based on machine vision [D]. Fujian Institute of Engineering, 2021.

Wu Chuang, Yu Dayong. Research on surface defect classification detection of mobile phone glass cover based on deep convolutional neural network [J]. Software Engineering, 2021, 24 (12): 6.

Jiang Qiyuan, Xie Jinxing,. Mathematical model. Version 3 [M]. Higher Education Press, 2003.

Cai Linlin. Model selection and parallelization of random forest [D]. Harbin Institute of Technology. 2012.

Ma Yunlong. RBF neural network prediction algorithm based on principal component analysis and its application [D]. Jilin University.

Zhang Kun, Shen Haibo, Zhang Hong, et al. Comprehensive evaluation method of node importance in complex network based on grey relational analysis [J]. 2022(4).

Zhou Jianhui, Meng Lei, Wang Lijuan, et al. Multiple correspondence analysis of pathogen distribution characteristics of fever with rash syndrome in Gansu Province from 2009 to 2019 [J]. China Public Health, 2022, 38 (3): 4.

Qi Mengsha, Li Wei, He Ping, et al. Obstacle survey and multiple correspondence analysis in pulmonary rehabilitation practice of chronic obstructive pulmonary disease [J]. Chinese Journal of Respiratory and Critical Care, 2021, 20 (2): 4.

Liu Zhen. Analysis on the Methods of Cultural Relics Identification. 2021.

Wang Chengyu, Tao Ying. Weathering of silicate glass. Journal of Silicate, 2003, 31 (1): 8.

Xue Xinju. Python-based K-means algorithm and its application. Science and Technology Horizon, 2018 (24): 2.