Classification of Fremont Ceramics Using a Neural Network

Summary

This is an abstract from the "SAA 2023: Individual Abstracts" session, at the 88th annual meeting of the Society for American Archaeology.

Ceramic classification is central to archaeological analysis, but without systematic and objective quantification, archaeologists cannot determine the definitive number of types or what they represent, despite decades of research. Recently archaeologists have applied machine learning models to improve the effectiveness of ceramic classification and extend the use of ceramics as a relative dating method for archaeological sites. To evaluate the efficacy of this approach, here we train a machine learning model on the Fremont type collection (n = 333) from the Basin-Plateau region of North America, and test how accurately a neural network can identify existing Fremont ceramic typologies. The neural network consistently distinguished between three of the six types of Fremont Gray Ware. These types are noted to have distinct decorative patterns. With larger sample sizes, the neural network would likely have increased success in identifying less distinctive patterns. This technique will be further applied to rock art from archaeological sites in Utah.

This Resource is Part of the Following Collections

• SAA 2023: Individual Abstracts   •
• Society for American Archaeology 88th Annual Meeting, Portland, OR (2023)

Keywords

Spatial Coverage

Record Identifiers

Abstract Id(s): 37217.0