Investigating Geospatial Arrangements of Stone Knapping at a Maya Lowland Site Using Random Forest Modeling

Summary

This is an abstract from the "Practice, Theory, and Ethics of Machine Learning in Archaeology" session, at the 90th annual meeting of the Society for American Archaeology.

The machine learning algorithm Random Forest has proven highly accurate in classifying archaeological soil and lithic microdebitage particles. Understanding this efficacy, this model was selected for implementation on soil samples collected from the market plaza of the Late Classic Maya site of Tzikin Tzakan. The ultimate aim of this effort is to discern the spatial arrangement of lithic microdebitage within this socioeconomic space, with the hope of contributing to a better understanding the interpersonal dynamics surrounding the practice of stone-knapping. Upon application of the Random Forest classification model to a subset of the soil samples collected at this site, nearly all samples tested contained lithic microdebitage particles. The highly frequent presence of lithic microdebitage within samples taken across the plaza indicates the strong likelihood of stone knapping practices on the plaza floor, with good likelihood that this practice existed within multiple areas of the plaza. The use of the Random Forest classifier on these samples thus points to the existence of nodes of stone knapping within this site’s plaza space.

This Resource is Part of the Following Collections

• Society for American Archaeology 90th Annual Meeting, Denver, CO (2025)   •
• Practice, Theory, and Ethics of Machine Learning in Archaeology

Keywords

Record Identifiers

Abstract Id(s): 51936