A classification model for ancient glass artefacts based on Fisher discriminant analysis and K-means clustering

Abstract

Classification is important for the study of ancient glass objects. Accurate identification and classification is important in presenting the cultural origins, heritage, and changes of glass artefacts. In order to accurately classify ancient glass artefacts, this paper first classifies ancient glass artefacts into two categories based on Fisher discriminant analysis, establishes a K-mean clustering model to subdivide the two types of artefacts, and further conducts sensitivity analysis to establish a classification model for ancient glass artefacts.