Semi-supervised machine learning approaches for predicting the chronology of archaeological sites: A case study of temples from medieval Angkor, Cambodia

A groundbreaking approach using semi-supervised machine learning has demonstrated the potential to accurately predict the chronology of medieval temples at Angkor, Cambodia, with an average error margin of 49-66 years. This method outperforms traditional statistical techniques, particularly in datasets with limited labeled data and incomplete attributes—a common challenge in archaeology. By leveraging a large dataset of temples with known architectural elements and artifacts (where less than 10% had confirmed dates), the study highlights how advanced computational power can address complex archaeological problems, offering a promising new tool for anthropological research.

Klassen, S., Weed, J., & Evans, D. (2018). Semi-supervised machine learning approaches for predicting the chronology of archaeological sites: A case study of temples from medieval Angkor, Cambodia. Plos One, 13(11), e0205649. https://doi.org/10.1371/journal.pone.0205649