Vizualizace hierarchických shlukovacích metod

Abstract

The aim of this bachelor thesis is to create web application showing and simulating a process of hierarchic aglomerative data clustering methods over two-dimensional data. The whole application acts as a tool instructing these methods. User has an option to choose one of three clustering strategies: the closest neighbour strategy, the furthermost neighbour strategy, and the middle object distance strategy. User has furthermore an option to set up animation parameters, input data, and clustering properties according which the animation is being generated. The animation procedure is then controlled by user. The application itself is an applet, quasi a special type of program written in Java programming language, which is being run in a web browser context.