Abstract

The article presents the first research on the elemental diversity of the prehistoric pottery obtained from the European Postglacial Plain, performed on a large scale (over 500 fragments of pottery were examined). A new analysis approach to the results of the elemental composition of the pottery was presented by using the unsupervised classification of chemical elements. This method is based on the advanced selection of features. The subset of elements that best describes the variability of the chemical composition of ceramics is searched, which allows to create the groups of ceramic artifacts corresponding to their origin or production technology. The correspondence between the existing locations and the clusters obtained using the Gaussian Mixture Model was controlled by using the entropy-based V-measure. Data analysis revealed that elemental composition of the pottery is more complex processes that simple provenance, possibly related to the emerging local pottery traditions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call