Application of multiple imputation using fuzzy archaeological data

Küçük Resim Yok



Dergi Başlığı

Dergi ISSN

Cilt Başlığı


Springer Verlag

Erişim Hakkı



This research offers a fuzzy approach to one of the missing data completion procedures, the Markov Chain Monte Carlo-based Multiple Imputation (MI) method. The implication of the method was processed through the data of 483 iron and bronze arrowheads dating back to Archaic Period (6th Century BC), gathered in Nif Mountain Excavation and Research Project located in İzmir, Western Anatolia region of Turkey, between the years 2006 and 2017. The main reason for selecting a data set based on the measurement results of archaeological gatherings is the uncertainty of the metric measurements due to the nature of the objects, such as corrosion, abrasion and decay. The fuzzification of MI as a Bayesian approach was applied through Fuzzy Bayesian Inference. The success of the offered approach is assessed by the assignment of right percentage values resulting from the application of the categorization model on the data set completed with Fuzzy MI, following the data completion process by the standard MI Method on the same data set. The categorization model above is based on the expert-based typology results, rooting from the visual similarities of the findings. © 2020, Springer Nature Switzerland AG.


International Conference on Intelligent and Fuzzy Systems, INFUS 2019 -- 23 July 2019 through 25 July 2019 -- -- 228529

Anahtar Kelimeler

Fully conditional specification, Fuzzy numbers, Missing value analysis, Multiple imputation


Advances in Intelligent Systems and Computing

WoS Q Değeri

Scopus Q Değeri