Predicting Flake Mass: A View from Machine Learning

ABSTRACT

Estimating flake mass based on remaining attributes bears an important relationship for the interpretation of lithic assemblages. Previous works have pointed out the relationship between flake attributes and prediction of flake mass. This study builds on previous works by using data from an experimental collection of flakes. Estimated mass was arrived at by generating a multiple linear regression model that combines several predictive variables. Variable selection for model training was carried out by using best subset selection, which evaluates all possible combinations of variables. Evaluation of the model was performed by computing common machine learning statistics along with estimated percentage error. Results make it possible to determine the best variables and estimate their relationships with flake mass. On the other hand, results also show that although the model is slightly biased and performs adequately, it has a limited inferential ability, especially when compared with other methods/indexes employed to estimate reduction.

KEYWORDS:

• Machine learning
• lithic analysis
• flake mass

Acknowledgements

The authors wish to thank the editor and the two anonymous reviewers for their comments and suggestions. This article is the result of the research projects “Como, Quien Y Donde?: Variabilidad De Comportamientos En La Captación Y Transformación De Los Recursos Liticos Dentro De Grupos Neandertales 2” (HAR2016-76760-C3-2-P) financed by Agencia Estatal de Investigación (AEI), Fondo Europeo de Desarrollo Regional (FEDER); and “En Los Limites De La Diversidad: Comportamiento Neandertal En El Centro Y Sur De La Penisula Iberica” (ID2019-103987GB-C33) financed by the Programa Estatal de Generación de Conocimiento y Fortalecimiento Científico y Tecnológico del Sistema de I + D+i y de I + D+i Orientada a los Retos de la Sociedad, del Plan Estatal de Investigación Científica y Técnica y de Innovación (2017–2020). Development of the experimentation and analysis of the materials were undertaken at the Laboratory of Experimental Archaeology (Universidad Autónoma de Madrid).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This article is the result of the research project HAR2016-76760-C3-2-P, “COMO, QUIEN Y DONDE?: VARIABILIDAD DE COMPORTAMIENTOS EN LA CAPTACION Y TRANSFORMACION DE LOS RECURSOS LITICOS DENTRO DE GRUPOS NEANDERTALES 2” financed by Agencia Estatal de Investigación (AEI) and Fondo Europeo de Desarrollo Regional (FEDER). Development of the experimentation and analysis of the materials were undertaken at the Laboratory of Experimental Archaeology (Universidad Autónoma de Madrid).

Notes on contributors

Guillermo Bustos-Pérez

Guillermo Bustos Pérez is PhD from the Universidad Autónoma of Madrid. He has dedicated the bulk of his PhD to the analysis of lithic technology from Middle Paleolithic Spanish sites and to the reduction process on retouched artifacts. Other focus of research are lithic taphonomy and application of programming languages for lithic analysis.

Javier Baena

Javier Baena is professor at Universidad Autónoma of Madrid. He has more than 30 years of experience on excavations that range from Acheulean to Chalcolithic, along as international projects at Olduvai and Dmanisi. Focus of research is on lithic analysis of Acheulean and Middle Paleolithic sites with special attention to experimental archaeology. He is also an expert knapper.