Title :
Radial basis function network based feature extraction for improvement the procedure of sourcing neolithic ceramics
Author :
Kvascev, G. ; Kvascev, M.G. ; Djurovic, Z.
Author_Institution :
Signals & Syst. Dept., Univ. of Belgrade, Belgrade, Serbia
Abstract :
Modern studies of a cultural heritage objects are increasingly multidisciplinary. A variety of analytical techniques supported by pattern recognition methods can help in answering about the origin, dating or authenticity. Results of sourcing ceramics from three Neolithic sites in Serbia are shown in this paper. The procedure based on radial basic function networks (RBFN) was employed in ceramic characterization. The results obtained were tested in previously established procedure for origin of production determination.
Keywords :
X-ray fluorescence analysis; archaeology; ceramics; feature extraction; history; radial basis function networks; signal classification; Neolithic site; RBFN; Serbia; analytical technique; archaeological ceramics; authenticity; ceramic characterization; ceramics classification; cultural heritage objects; dating; feature extraction; neolithic ceramics sourcing; origin; pEDXRF; pattern recognition method; portable energy dispersive X-ray fluorescence; production determination; radial basic function network; radial basis function network; Ceramics; Chemical elements; Classification algorithms; Decision making; Feature extraction; Training; Vectors; Classification; Dimension reduction; Feature extraction; Pattern recognition; Radial basis function network;
Conference_Titel :
Neural Network Applications in Electrical Engineering (NEUREL), 2012 11th Symposium on
Conference_Location :
Belgrade
Print_ISBN :
978-1-4673-1569-2
DOI :
10.1109/NEUREL.2012.6419973
