Predictive Model for Chinese Excavated Glass Based on Least Squares Method and BP-Neural Network
Jiantao Song · Atlantis Highlights in Computer Sciences/Atlantis highlights in computer sciences · 2023
The ancient glass excavated in China proves that at least in the Warring States period more than 2,000 years ago, it was already possible to manufacture glass products with exquisite patterns.The antique Chinese lead-barium glass and potassium glass represent China's indigenous glass technology system, and their study has significantly contributed to the history of Chinese science and technology.To this end, this paper provides a physical examination of a group of more representative ancient Chinese lead-barium glass and potassium glass.Old glass is susceptible to weathering by composition and environment.Environmental factors mainly refer to the temperature, humidity, and time when storing glass.Therefore, the weathering products on the glass surface are determined by the glass composition, temperature, humidity, time, and atmosphere at the weathering time.The ratio of the chemical composition of the weathered glass will change and thus affect the judgment of the glass type.Therefore, this study developed partial least squares regression and PB neural network models to predict the chemical composition of lead-barium glass and potassium glass from a group of Chinese excavations.