Data Analysis II - full-time study

Network Construction

Methods of network construction from vector data based on weighting (similarity) of objects and different approaches of edges selection.

In the seminar, a selected method is implemented and applied to the analysis of suitable datasets.

Data Structures for Network Representation

Advanced Network Models I

Community network models, triadic closure as a basic principle generating community network structure.

Implementation of a model using the triadic closure (Bianconi et al.), experiments, visualization.

Advanced Network Models II

Network models with preferential attachment, evolving networks. Preferential attachment as a principle and as a result of another principle.

Implementation of models generating preferential attachment. Experiments and visualization.

Link Prediction

Link prediction in networks. Different approaches, methods based on local similarity.

Using and comparing different methods based on similarity and on the analysis of common neighbors.