Past Movement and New Models: Reconstructing Past Mobility in the Absaroka Mountains by Applying Bayesian Neural Networks Towards Refining Trace Element Modeling

Summary

This is an abstract from the "Machine-Learning Approaches to Studying Ancient Human-Environmental Interactions" session, at the 90th annual meeting of the Society for American Archaeology.

Sourcing lithic raw materials in North America has become increasingly valuable for understanding past human behavior. However, the process often faces challenges due to monetary costs and the need to remove materials from their original landscape. Refining pXRF obsidian sourcing methods can help mitigate these issues. By utilizing a large dataset from the GRSLE project, which includes obsidian sourced through eXRF and pXRF scans, we developed a multinomial model using Bayesian neural networks to predict obsidian sources from pXRF scans. The Bugas-Holding site, included in this analysis, highlights the benefits of understanding within-site patterning of obsidian. The application of machine learning techniques not only enhances our knowledge of the past but also improves the ability of archaeologists to record data for future research.

This Resource is Part of the Following Collections

• Society for American Archaeology 90th Annual Meeting, Denver, CO (2025)   •
• Machine-Learning Approaches to Studying Ancient Human-Environmental Interactions

Keywords

Record Identifiers

Abstract Id(s): 51330