Using Supervised Machine Learning for Modelling Early Neolithic Survival Probability: a Bayesian Networks approach

Abstract

The present study explores the application of the supervised machine learning approach to archaeological studies. The goal is to develop a supervised machine learning model based on Bayesian networks with a dual purpose: on the one hand, it will be a predictive model to assess the probability of survival (equivalently, the risk of disappearance) of an early agropastoral community. On the other, the model will allow us to understand the importance of and the relationships between factors affecting a subsistence model of early farming (ca. 7500/7000 - 5100 cal. BP approximately) in the northeast Iberian Peninsula. Our work embodies multiple factors relevant to understanding social and economic decision-making in prehistoric communities, associating archaeological observations with theoretical parameters with the goal of offering new insights into socioeconomic and technological systems of early Neolithic societies.