| Abstract [eng] |
The following thesis is about methods of spatial analysis for archaeological research. The main aim of this work is to identify appropriate methods of spatial analysis for these tasks: measuring areas of archaeological cultures and determining archaeological sensitivity in selected space (predictive modeling), combine them into models and implement them in geographic information systems (GIS). First - existing literature on the topic was analysed, the usual approaches to these and related problems were determined. Consequently during the next phase, methods that aren‘t yet very popular in archaeological modeling were selected: archaeological culture area measurement based on „Xtent“ model in vector data model and ensemble machine learning algorithm – random forest, for determining archaeological sensitivity of selected area. Finally to implement these methods, standard GIS functionality was extended by using statistical analysis system „R“ and Python programming language. The results show, that selection of methods for these tasks can be considered appropriate. The models created during this research could be used as scientific purpose analytical tool for archaeologists who are concerned with environment factors that influenced prehistoric societies and extent of their cultures in space. Nonetheles, a predictive model could also be used in construction sector – as a tool for planning, to identify increased risk areas, where archaeological sensitivity is high, and small changes in planning could significantly reduce the risk of mandatory, expensive and time consuming archaeological excavations. |
