L’article “Unsupervised model-based clustering for typological classification of Middle Bronze Age flanged axes” par J. Wilczek et al. publié dans Journal of Archaological Science : Reports

photo2

Maréva Gabillot et Josef Wilczek examinent la documentation archéologique

pdfL’article “Unsupervised model-based clustering for typological classification of Middle Bronze Age flanged axes” par J. Wilczek, F. Monna, M. Gabillot, N. Navarro, L. Rusch, C. Chateau (2015) vient d’être publié dans Journal of Archaeological Science : Reports. [pdf]

Josef Wilczek (josef.wilczek@hotmail.com) vient de publier une série de résultats acquis durant son doctorat. Il s’agit de proposer une méthode de classification automatique pour les haches à rebord datant de l’Age du Bronze Moyen, qui jusque là était traitées en seulement deux groupes génériques le groupe atlantique et le groupe oriental. L’application des techniques les plus récentes de morphométrie géométrique permet d’isoler six groupes de formes. Ces derniers sont validés spatialement grâce une analyse spatiale entreprise sur le lieu de découverte des objets.

Abstract: The classification of Western European flanged axes dating to the Middle Bronze Age (1650–1350 BC) is very complex. Many types of axe have been identified, some of which have numerous variant forms. In the current French terminology, all axes are divided into two generic groups: namely “Atlantic” (Atlantique) and “Eastern” (Orientale). Each of these generic groups, however, is highly polymorphic, so that it is often very difficult for the operator to classify individual axes with absolute confidence and certainty. In order to overcome such problems, a new shape classification is proposed, using morphometric analysis (Elliptic Fourier Analysis) followed by unsupervised model-based clustering and discriminant analysis, both based on Gaussian mixture modelling. Together, these methods produce a clearer pattern, which is independently validated by the spatial distribution of the findings, and multinomial scan statistics. This approach is fast, reproducible, and operator-independent, allowing artefacts of unknown membership to be classified rapidly. The method is designed to be amendable by the introduction of new artefacts, in the light of future discoveries. This method can be adapted to suit many other archaeological artefacts, providing information about the material, social and cultural relations of ancient populations.

Comments are closed.