Application of machine learning algorithms for computational palaeography

OData support
Supervisor:
Dr. Hosszú Gábor
Department of Electron Devices

The main objective of machine learning is to develop such systems, that can learn from previous experiences. This means that these systems can recognize patterns based on provided data samples, with or without human interference. These methods have widespread utilization capabilities, they can be used for various data mining tasks, in areas such as computer engineering, economics or even psychology. The main cause of the rapid development of machine learning is the huge increase of the amount of data we need to process. It’s enough to look at the rapidly developing face-, voice- and handwriting recognition softwares to understand the need for fast and reliable machine learning methods. My purpose was to examine such methods that can help in paleographical researches, where these methods were rarely used before. The methods examined by me consist cluster analisys, classification, dimension reduction(multidimensional scaling, principal component analisys) and various other methods such as variance analisys and correlation analisys.

Downloads

Please sign in to download the files of this thesis.