Making Machine Learning More Accessible and Useful in Archaeology: Insights from Chronology Building

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.

While machine learning is beginning to appear in the archaeological literature, most archaeologists remain unfamiliar with this potentially useful analytical toolkit. Over the past decade, we have been exploring machine learning as a robust way to help address a significant challenge of the archaeological record: chronologically unmixing palimpsest lithic surface collections. Machine learning provides a reproducible, quantitative, and defensible method for chronological unmixing. Our experience also has given us more general insights to share that we hope will encourage others to make use of this approach. We discuss out how machine learning is less of a novel and mysterious method that many colleagues believe; that in many cases machine learning can be used with computers already accessible to most archaeologists; best practices in applying machine learning also promote robust and reproducible science; while machine learning is not relevant for all analyses of archaeological data it is very useful for many of the problems that archaeologists often wrestle with; and if the field of archaeology is to benefit from machine learning, it is imperative that archaeologists share their data in open, useable, and ethical ways. We use our work on chronological unmixing to exemplify these points.

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): 51930