Soft Computing Techniques for Classification of Bronze Age Axes

In this paper, the application of Multiple Classifier Systems and Soft Computing techniques to the classification of Bronze Age axes found in Italian territory is shown. The methodology used from feature extraction to classification is detailed. The results are obtained by using a data set of 85 axes, with training accomplished by bootstrapping the data. The system has been tested on new axes to be classified and validated on an artificial data set generated following the covariance matrices of the original archaeological data.