Discovering geo-magnetic anomalies: a clustering-based approach

The paper investigates the application of clustering algorithms to data obtained by magnetometry in the places of archaeological excavations. The goal was to enable automatic detection in order to find possible anomalies in the data that could indicate the presence of potential archeological sites. Analysis of the results of following experiments with clustering algorithms in presented: k-means, c-means and DBScan. The algorithms were run on a number of datasets, and visual assessment by experts confirmed the high level of accuracy of the results.